Metformin for Treatment of Overweight Induced by Atypical Antipsychotic Medication in Young People With Autism Spectrum Disorder A Randomized Clinical Trial

By Evdokia Anagnostou, MD1; Michael G. Aman, PhD; Benjamin L. Handen, PhD, et al.

ABSTRACT

Importance Atypical antipsychotic medications are indicated for the treatment of irritability and agitation symptoms in children with autism spectrum disorder (ASD). Unfortunately, these medications are associated with weight gain and metabolic complications that are especially troubling in children and with long-term use.

Objective To evaluate the efficacy of metformin for weight gain associated with atypical antipsychotic medications in children and adolescents with ASD (defined in the protocol as DSM-IV diagnosis of autistic disorder, Asperger disorder, or pervasive developmental disorder not otherwise specified), aged 6 to 17 years.

Design, Setting, and Participants A 16-week, double-blind, placebo-controlled, randomized clinical trial was conducted at 4 centers in Toronto, Ontario, Canada; Columbus, Ohio; Pittsburgh, Pennsylvania; and Nashville, Tennessee. In all, 209 potential participants were screened by telephone, 69 individuals provided consent, and 61 participants were randomized to receive metformin or placebo between April 26, 2013, and June 24, 2015.

Interventions Metformin or matching placebo titrated up to 500 mg twice daily for children aged 6 to 9 years and 850 mg twice daily for those 10 to 17 years.

Main Outcomes and Measures The primary outcome measure was change in body mass index (BMI) z score during 16 weeks of treatment. Secondary outcomes included changes in additional body composition and metabolic variables. Safety, tolerability, and efficacy analyses all used a modified intent-to-treat sample comprising all participants who received at least 1 dose of medication.

Results Of the 61 randomized participants, 60 participants initiated treatment (45 [75%] male; mean [SD] age, 12.8 [2.7] years). Metformin reduced BMI z scores from baseline to week 16 significantly more than placebo (difference in 16-week change scores vs placebo, −0.10 [95% CI, −0.16 to −0.04]; P = .003). Statistically significant improvements were also noted in secondary body composition measures (raw BMI, −0.95 [95% CI, −1.46 to −0.45] and raw weight, −2.73 [95% CI, −4.04 to −1.43]) but not in metabolic variables. Overall, metformin was well tolerated. Five participants in the metformin group discontinued treatment owing to adverse events (agitation, 4; sedation, 1). Participants receiving metformin vs placebo experienced gastrointestinal adverse events during a significantly higher percentage of treatment days (25.1% vs 6.8%; P = .005).

Conclusions and Relevance Metformin may be effective in decreasing weight gain associated with atypical antipsychotic use and is well tolerated by children and adolescents with ASD.

Trial Registration clinicaltrials.gov Identifier: NCT01825798

INTRODUCTION

The atypical antipsychotic medications risperidone and aripiprazole are the only treatments approved by the US Food and Drug Administration for use in autism spectrum disorder (ASD). Both medications improve irritability and agitation symptoms in children with ASD as young as 5 (risperidone) or 6 (aripiprazole) years.1- 6 Unfortunately, these medications also cause weight gain,7,8 and greater cumulative exposure is also associated with increased diabetes risk.9,10 Although weight gain may be reversible,11,12 stopping these medicines often leads to relapse of irritability and agitation.13 To our knowledge, no controlled studies have tested options for combating this weight gain in children with ASD.

Metformin hydrochloride, a biguanide drug approved for treatment of type 2 diabetes, increases insulin sensitivity and decreases intestinal glucose absorption and hepatic glucose production. Metformin has been studied for treatment of weight gain in the absence of diabetes in typically developing children.14,15 In adults, metformin can stop or reverse weight gain associated with atypical antipsychotics.16- 18 Metformin significantly reduced body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) in a randomized clinical trial of 39 children aged 10 to 17 years who were receiving atypical antipsychotic medications,19 similar to 2 open-label trials.20,21 Results of a smaller randomized clinical trial failed to reach statistical significance.22 Adverse events (AEs) in these studies were typically mild and largely confined to gastrointestinal (GI) symptoms, including diarrhea and nausea.14- 21

Because atypical antipsychotic medications are often used in children with ASD and the mechanisms of weight gain are thought to include insulin resistance, 23 we sought to assess the safety, tolerability, and efficacy of metformin to decrease weight gain associated with the use of atypical antipsychotic medication in children with ASD. Unlike previous trials, we studied a younger age group that is at greater cumulative risk due to extended antipsychotic exposure. In addition, we focused exclusively on children with ASD who are less able to communicate potential AEs, thus meriting a separate assessment of safety and tolerability. We were concerned that GI AEs due to metformin15 could overlap with GI symptoms that are already common in ASD,24 potentially increasing irritability in children who cannot readily communicate discomfort.

Our specific hypotheses were that, at the end of 16 weeks, (1) metformin would ameliorate weight gain compared with placebo (primary outcome measure, BMI z score; secondary outcome measures, BMI, weight z score, and weight) and (2) metformin would be well tolerated based on AE rates and lack of worsening of the irritability/agitation targeted by atypical antipsychotic treatment.

Key Points
Question What is the effect of metformin hydrochloride on weight gain associated with the use of atypical antipsychotics in children and adolescents with autism spectrum disorders?

Findings In this randomized clinical trial of 60 participants, metformin significantly reduced weight gain compared with placebo. Overall, metformin was well tolerated during 16 weeks of treatment.

Meaning In children and youth with autism spectrum disorder who are receiving atypical antipsychotics, metformin may be effective in decreasing weight gain.

METHODS

Study Design and Participants
This was a 16-week, randomized clinical trial testing the safety, tolerability, and efficacy of a liquid formulation of metformin hydrochloride (Riomet) in children and adolescents with ASD (Figure 1). The trial was conducted from April 26, 2013, to June 24, 2015. Participants were recruited through 4 academic sites participating in the Autism Speaks Autism Treatment Network. This trial was approved by the institutional review boards at each study site, and all participants or their legal guardians signed institutionally approved informed consent or assent forms according to the Helsinki agreement.25 The full trial protocol is given in Supplement 1. Participants received financial compensation.

Figure 1.
CONSORT Flow Diagram

Image not available.
View Large | Save Figure | Download Slide (.ppt)
Inclusion criteria were age 6 to 17 years, 4 months; diagnosis of ASD (defined in the protocol as DSM-IV diagnosis of autistic disorder, Asperger disorder, or pervasive developmental disorder not otherwise specified) based on the Autism Diagnostic Observation Schedule26 and DSM-IV clinical interview27; minimum of 1 month of therapy with a stable dose of an atypical antipsychotic with no plans to change; and a documented 7% or more increase in BMI since starting an atypical antipsychotic (within the past 12 months), or, if the BMI was in the 85th percentile or higher, a greater than 5% body weight increase per year since starting the medication, as documented by previous weight records.

Exclusion criteria were a history of intolerable AEs with metformin, previous use of metformin of sufficient dose and duration to determine response status, any serious medical illness requiring treatment or increasing the risk of lactic acidosis, use of medication with unacceptable interactions with metformin, planned procedure requiring contrast medium, pregnancy at screening, current use of medication for target symptoms of appetite or weight loss, or inability to tolerate blood work.

Screening Assessments
Diagnostic assessment included the Autism Diagnostic Observation Schedule26 and DSM-IV interview27; cognitive assessment was performed using the Stanford-Binet, 5th Edition, Abbreviated Battery28 or Mullen Scales of Early Learning29; and behavior was evaluated by the caregiver-completed Aberrant Behavior Checklist.30 A comprehensive medical and developmental assessment included vital signs, height, weight, physical examination, review of medical and psychiatric history including Safety Monitoring Uniform Report Form,31,32 review of concomitant therapies, and clinical laboratory tests (liver function, metabolic panel, serum lactate, fasting glucose, hemoglobin A1c, insulin, vitamin B12, lipid panel, and complete blood cell count). Urine or blood pregnancy testing was conducted for girls of child-bearing potential.

Study Intervention
All participants and caregivers received brief counseling and informational handouts regarding diet and exercise before they began the trial. Metformin hydrochloride (Riomet) and matching placebo were donated by Ranbaxy Pharmaceuticals Ltd. Metformin was dispensed as a liquid suspension of 100 mg/mL, with placebo matching the appearance, smell, and taste of metformin. For 6- to 9-year-old children, the dosage was started at 250 mg with their evening meal for 1 week, followed by the addition of 250 mg at breakfast for 1 week. At the week 2 visit, if the previous dose had been well tolerated, it was increased to 500 mg twice daily. For 10- to 17-year-old participants, treatment was started following the same schedule as used with the children, but at the week 4 visit, if the drug had been well tolerated, the dose was increased to 850 mg twice daily. The dosing strategy was based on previous pediatric trials of metformin.15,19 If AEs occurred, the study physician (including E.A., K.B.S., J.A.H., L.E.A., L.C., D.M., R.T., J.K., and J.V.-V.) had the option to decrease the dose in multiples of 50 mg and rechallenge once the AE was resolved, with higher doses administered in multiples of 50-mg units.

All treatment was double-blinded. Randomization was conducted by the data coordinating center via permuted blocks stratified by age group (6-9 vs 10-17 years) and site using random block sizes of 2 and 4. An autogenerated email with randomization identification was sent to the site once a participant successfully enrolled. Only the investigational pharmacy at each site and unblinded staff at the data coordinating center had access to the treatment assignment.

Primary Outcome
The primary efficacy measure was change from baseline to week 16 in body mass index (BMI) z score calculated from the 2000 Centers for Disease Control and Prevention (http://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm) age- and sex-normed growth charts at week 16. Height and weight were measured at baseline and 2, 4, 8, 12, and 16 weeks after treatment initiation.

Secondary Outcomes
Secondary outcomes included (1) changes in other body variables (absolute and relative change in weight, absolute BMI, and abdominal and hip circumference) and (2) changes in fasting metabolic variables (eTable 1 and eTable 2 in Supplement 2). The Homeostasis Model Assessment of Insulin Resistance33 was used to integrate fasting glucose and insulin values in a composite measure of metabolic risk.

Safety Assessments
Safety was assessed with the Safety Monitoring Uniform Report Form,31,32 which was slightly modified to assess rare AEs previously reported with metformin,14- 22 which was administered by a clinician at every visit. Clinical laboratory tests were repeated at week 16.

Given concern that children with ASD may be unable to report AEs verbally and could manifest physical discomfort with increased irritability/agitation,34- 36 we examined the effect of metformin vs placebo on irritability/agitation by caregiver ratings on the Aberrant Behavior Checklist Irritability subscale.29,37 The Clinical Global Impression Scale–Improvement,38 based on change in global symptoms, including behavioral and physical functioning, was also administered at every visit as a measure of tolerability.31

Statistical Analysis
A sample size of 90 participants was originally planned, assuming a pooled, among-person SD of 0.6 for a 16-week change in BMI z score. Interim blinded analysis of change in BMI z scores indicated an effective SD of only 0.16, and the sample size was reduced to 60 for 80% power to detect a difference of 0.12, which would be a difference in weight change of approximately 3.0 kg in our sample. The interim analysis was of blinded data analyzed as a single sample only and thus did not affect our type I error rate.39

Safety, tolerability, and efficacy analyses used a modified intent-to-treat sample comprising all participants who received at least 1 dose of medication. The primary efficacy end point was 16-week change from baseline in BMI z scores. All participants in the modified intent-to-treat sample were included in the analysis. Secondary analysis included all participant randomized, including 1 individual who did not receive medication (intent to treat).

The proportion of participants experiencing AEs classified by the MedDRA (http://www.meddra.org/) system organ class and preferred term was compared between treatments by using the Fisher exact test with estimated odds ratios and exact 95% confidence intervals. Effects of metformin on primary and secondary efficacy outcomes were estimated from shared-baseline, random-slope, linear mixed models with fixed effects of stratum × visit (categorical) and stratum × treatment × postbaseline visit interaction, and random participant-specific intercepts and slopes with unstructured covariance. The mean model was unstructured in time, whereas the covariance model assumed participant-specific linear deviations from the means. The shared-baseline assumption, enforced by omitting a treatment main-effect term, reflected the true state of the population before randomization and adjusts for chance differences at baseline.40 Effects of treatment assignment on 16-week change were estimated by linear contrasts of the baseline and week 16 least-square means using the observe stratum frequencies. Effect sizes (ESs) for treatment differences were calculated relative to the pooled SD for 16-week change among participants who completed the trial. The primary end point was tested at 2-tailed α = .05. The secondary end points and analyses were considered exploratory, without adjustment for multiple comparisons. To assess whether GI AEs mediated benefit from metformin, we calculated the number of days that participants experienced any of the following symptoms: abdominal discomfort or pain, diarrhea, eructation, flatulence, nausea, vomiting, and gastroenteritis. We tested whether total duration of GI AEs mediated the treatment effect by using a causal model for direct and indirect effects.41 The total natural indirect effect for GI AEs was taken as a measure of mediation.

RESULTS

Participants
Sixty-one participants were randomized and 60 initiated treatment, 28 (7 [25%] female) received metformin and 32 (8 [25%] female) received placebo, and the mean (SD) age was 12.8 (2.7) years (Figure 1, Table 1, and eTable 3 in Supplement 2). One participant randomized to receive metformin refused liquid medication and therefore withdrew without ever taking treatment. This participant was included only in the secondary ITT analysis.

Table 1. Baseline Characteristics
Image not available.
View Large | Save Table | Download Slide (.ppt)
At baseline, the placebo group had a higher IQ than the metformin group (mean [SD], 84 [20] vs 69 [25]) and was receiving more psychotropic medications (53% vs 36% were taking ≥3 medications) (Table 1 and eTable 4 in Supplement 2).

In children aged 6 to 9 years, the mean final dose in both the metformin and placebo groups was 1000 mg/d. In those aged 10 to 17 years, the mean final dose in the metformin group was 1587 mg/d and in the placebo group, 1674 mg/d (P = .24). Adherence was excellent in both groups (96% of doses taken in both groups, with ranges of 75%-100% in the metformin group and 85%-100% in the placebo group).

Primary Outcome
To account for anticipated growth and differences in typical BMI across childhood, we used changes in BMI standardized to age- and sex-based norms (z scores). Metformin was superior to placebo in reducing weight gain associated with atypical antipsychotics, as assessed by change from baseline to week 16 BMI z scores (metformin vs placebo difference, −0.010 [95% CI, −0.16 to −0.04]; ES, 0.82; P = .003) (Table 2, Figure 2, and pretreatment and posttreatment values in eTable 5 in Supplement 2). Participants receiving placebo experienced no change over 16 weeks in BMI z scores (0.02 [95% CI, −0.03 to 0.06]); however, those receiving metformin experienced a substantial reduction compared with baseline (−0.08 [95% CI, −0.13 to −0.04]). Three of 28 participants (11%) in the metformin group experienced 8% to 9% declines in BMI. No other participants experienced more than a 5% decline in BMI during the 16-week treatment period. Benefit from metformin did not differ significantly by study site (P = .91) (eTable 6 in Supplement 2).

Table 2. Effect of Metformin vs Placebo on Primary/Secondary Measures: Anthropometric Measurements and Irritabilitya
Image not available.
View Large | Save Table | Download Slide (.ppt)
Figure 2.
Metformin Effect on Body Mass Index (BMI) z Score and Weight Change

Estimates of means for the treatment and placebo groups at each visit. Error bars indicate 95% CIs.

aSignificant difference at each visit.

Image not available.
View Large | Save Figure | Download Slide (.ppt)
Secondary Outcomes
A similar pattern of differences between the treatment and placebo group was observed when examining weight z scores (ES, 1.04), BMI (ES, 1.01), weight (difference, 2.7 kg; ES, 1.13), BMI percentiles (ES, 0.47), and weight percentiles (ES, 0.65) (Table 2 and Figure 2). Baseline BMI was positively but weakly associated with percentage change in BMI (r = 0.11; P = .39) (eFigure in Supplement 2). No significant differences were noted in changes of any of the metabolic variables studied (eTable 5 in Supplement 2). Intent-to-treat analysis, including 1 participant who never received the study medication, did not lead to meaningful changes in the results (eTable 7 in Supplement 2).

Safety
There was no significant difference between the groups in change from baseline to week 16 on the Aberrant Behavior Checklist Irritability subscale (ES, 0.01; P = .94) or the Clinical Global Impression Scale Improvement Scale (ES, 0.11; P = .91). There was 1 serious AE. A participant with a history of hospital admissions for aggression who was randomized to receive placebo was admitted for aggression. The serious AE was determined to be unrelated to the study drug. One other participant receiving placebo was lost to follow-up when the participant withdrew consent after the dose of the atypical antipsychotic medication was increased following an outburst at school. Five participants in the metformin group discontinued treatment: 4 owing to agitation (determined to be possibly related to the study drug in 3 and unrelated in 1, at weeks 1, 5, and 8, with 2 at week 8) and 1 owing to sedation (at week 10).

Adverse events occurring in more than 5% of participants in either group (>1 participant) are listed in Table 3. Overall, abnormal feces was the only statistically significant difference in AEs between the groups (4 of 28 [14%] receiving metformin vs 0 of 32 receiving placebo; P = .04). There was a nonsignificant increase in GI AEs in the metformin group compared with the placebo group (23 [82%] receiving metformin vs 19 [59%] receiving placebo; P = .09) (Table 3). Participants receiving metformin experienced GI AEs during a significantly higher percentage of treatment days (25.1% vs 6.8%; P = .005).

Table 3. Adverse Eventsa
Image not available.
View Large | Save Table | Download Slide (.ppt)
Blinding Assessment
Both clinicians and participants were asked to guess whether the participant was taking metformin or placebo at the end of 16 weeks. The rate of guessing participant assignment correctly did not differ significantly between the metformin and placebo groups for either clinician (metformin: guessed metformin, 8%; placebo, 15%; uncertain, 77%; placebo: guessed metformin, 10%; placebo, 24%; uncertain, 66%) or participant (metformin: guessed metformin, 32%, placebo, 36%, uncertain, 32%; placebo: guessed metformin, 14%, placebo, 61%, uncertain, 25%) guesses. Among caregivers who guessed (excluding those who were uncertain), caregivers of placebo participants more often guessed correctly (17 of 21 [81%]; odds ratio, 3.8 [95% CI, 0.7 to 2.15]; P = .09). Correct guessing by caregivers did not predict change in the primary outcome measure (16-week change in BMI z score among placebo correct guessers, −0.001; among placebo incorrect guessers, −0.025; difference, 0.023 [95% CI, −0.08 to 0.12]; P = .65).

Post Hoc Analysis
Baseline variables differing substantially between groups included aripiprazole use (metformin, 7 [25%]; placebo, 16 [50%]), risperidone use (metformin, 19 [68%]; placebo, 12 [38%]), and mean [SD] full-scale IQ (metformin, 69 [25]; placebo, 84 [20]). There was no evidence that any of these findings confounded the association between randomized treatment assignment and changes in BMI z scores: use of aripiprazole (interaction ES, 0.3; P = .61), use of risperidone (interaction ES, 0.26; P = .64), age (interaction ES, 0.001 per year; P = .91), full-scale IQ (interaction ES, 0.005 per unit; P = .71), and number of psychotropic medications (interaction ES, 0.09 per medication; P = .68).

We were concerned that the apparent benefit of metformin might only be the result of increased GI discomfort. We found, however, that participants experiencing more days with GI symptoms experienced insignificantly greater (rather than smaller) increases in BMI and weight. The total natural indirect effect mediated by GI AEs were gains of 0.007 for BMI z scores (95% CI, −0.027 to 0.042; P = .68) and 0.11 kg for weight (95% CI, −0.57 to 0.80; P = .75).

DISCUSSION

Atypical antipsychotic medications are US Food and Drug Administration–approved for use in children at a younger age in ASD than in any other condition; however, to our knowledge, no previous studies have examined treatment or prevention of weight gain in children with ASD who receive these medications. Metformin was superior to placebo in managing established weight gain associated with atypical antipsychotics in this population. Children receiving placebo continued to gain weight during the study, as expected for age (BMI z score increase negligible), whereas BMI z scores decreased in the metformin group (Figure 2). The large effect size of BMI z score changes in this study is comparable to that observed in the Klein et al19 pediatric pilot trial of metformin in children receiving atypical antipsychotics. The benefits from treatment in our sample were not clear until after 8 weeks of treatment (Figure 2), which was also consistent with previous pediatric RCTs15,19 of metformin targeting weight gain in children and youth. Other strategies to reduce weight gain associated with atypical antipsychotic use in adults include sibutramine hydrochloride and topiramate, but the AE profiles of these drugs are not trivial.42 In children, stimulants failed to reduce weight gain associated with atypical antipsychotics.43 In addition, nutritional counseling has previously shown little benefit for children and adolescents receiving atypical antipsychotic medications,19 and food selectivity may make this particularly challenging in the ASD population.26 However, metformin was not only effective, it was well tolerated in our participants. Although there was nonsignificant increase in GI distress in the metformin group, GI AEs were not responsible for any treatment discontinuation, and we found no evidence that the benefit on BMI was due to GI AEs.

There are some important limitations to this study. First, the length of metformin treatment may have been inadequate to capture metabolic benefits. Furthermore, the length of the study was too short to evaluate whether initial improvements will be maintained. The small sample size yielded too little precision in our estimates to discount possible effects for metabolic and behavioral outcomes. In addition, although we had reliable information on the length of time that participants were stable on their current dose, we did not have reliable information on the length of overall exposure to antipsychotic medications. We also did not evaluate the benefit of lifestyle modifications, although brief counseling regarding diet and exercise was provided to all participants. Behavioral interventions targeting weight in youth with ASD have not been tested. Combined programs that compared metformin with placebo with all participants receiving vigorous and continuous lifestyle supports might produce different results. Finally, this study addressed only the benefit of metformin added to an atypical antipsychotic after children had already gained significant weight. The trial did not address the question of whether coadministration of metformin at the onset of atypical antipsychotic use prevents initial weight gain, which is a question for future research.

Conflict of Interest Disclosures: Dr Anagnostou has received consultation fees and served on advisory boards for Roche; industry funding from SynapDx and Aventis; royalties from APPI, Springer International Publishing; and other funding from Canadian Institutes of Health Research, Ontario Brain Institute, Department of Defense, Autism Speaks, National Centers of Excellence, and Physician Services Incorporated. Dr Aman has received consultation fees and research contracts, served on advisory boards, or provided investigator training for Bristol-Myers-Squibb, CogState, Ltd, Confluence Pharmaceutica, Coronado Biosciences, Forest Research, Roche, Janssen Pharmaceuticals–Johnson and Johnson, Lumos Pharma, MedAvante, Inc, Novartis, ProPhase LLC, and Supernus Pharmaceuticals. Dr Handen has received funding from Lilly, Curemark, and Roche, National Institute of Mental Health, National Institute of Aging, and Autism Speaks. Dr Sanders has received research funding from Roche, Forest, Curemark, Sunovion, and Stemina; and other funding from Autism Speaks, Health Resources and Services Administration (HRSA), and National Institute for Child Health and Human Development. Dr Hollway has received funding from Roche, Supernus, Forest Research Institute, Sunovion, Young Living Essential Oils, and Autism Speaks. Dr Arnold has consulted for or served on advisory boards of Arbor, Gowlings, Ironshore, Neuropharm, Novartis, Noven, Organon, Otsuka, Pfizer, Roche, Seaside Therapeutics, Sigma Tau, Shire, Tris Pharma, and Waypoint; has received research funding from Curemark, Forest, Lilly, Neuropharm, Novartis, Noven, Shire, Supernus, and Young Living; other funding from National Institutes of Health and Autism Speaks; and travel support from Noven. Dr Hellings has served as an advisor for Abbott Laboratories and received research funding from Roche, Sunovion, Supernus, Young Living Essential Oils, Forest, and Shire. Dr Tumuluru has received funding from Lilly, Curemark, Roche, National Institute of Mental Health, and NIA. Dr Kettel has received funding from Autism Speaks. Dr Macklin is a Data Safety Monitoring Board member of Acorda Therapeutics and Shire Human Genetic Therapies and has received grant support from Autism Speaks, Adolph Coors Foundation, ALS Association, ALS Therapy Development Initiative, ALS Therapy Alliance, Biotie Therapies, Michael J. Fox Foundation, Muscular Dystrophy Association, HRSA, and National Institutes of Health. Dr Veenstra-VanderWeele has consulted for or served on the advisory boards of Roche, Novartis, SynapDx; received research funding from Roche, Novartis, Seaside Therapeutics, Forest, and SynapDx; and has received other funding from the Sackler Foundation, New York Collaborates for Autism, Autism Speaks, Landreth Family Discovery Grant, Vanderbilt University, Columbia University, National Institute of Mental Health, National Institute for Child Health and Human Development, Autism Speaks, HRSA, Agency for Health Research and Quality. No other disclosures were reported.

Funding/Support: This project is/was funded by the HRSA of the US Department of Health and Human Services (HHS) under cooperative agreement grant UA3 MC11054–Autism Intervention Research Network on Physical Health. Ranbaxy Laboratories Ltd donated both metformin and placebo for the purposes of this study.

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: This information or content and conclusions are those of the authors and should not be construed as the official position or policy of, nor should any endorsements be inferred by, HRSA, HHS, or the US government.

Additional Information: This work was conducted through the Autism Speaks Autism Treatment Network serving as the Autism Intervention Research Network on Physical Health.

Additional Contributions: Other contributing DSMB members included Christopher J. Kratochvil, MD (UNeHealth, University of Nebraska Medical Center), Wesley K. Thompson, PhD (Department of Psychiatry, University of California at San Diego, La Jolla), Stephen Cook, MD, MPH (Department of Pediatrics, University of Rochester Medical Center), and Walter J. Meyer, MD (Department of Psychiatry and Behavioral Sciences, University of Texas Medical Branch, Galveston). This service was compensated.

REFERENCES

1
McCracken JT, McGough J, Shah B, et al; Research Units on Pediatric Psychopharmacology Autism Network. Risperidone in children with autism and serious behavioral problems. N Engl J Med. 2002;347(5):314-321.
PubMed | Link to Article
2
McDougle CJ, Scahill L, Aman MG, et al. Risperidone for the core symptom domains of autism: results from the study by the autism network of the research units on pediatric psychopharmacology. Am J Psychiatry. 2005;162(6):1142-1148.
PubMed | Link to Article
3
Marcus RN, Owen R, Kamen L, et al. A placebo-controlled, fixed-dose study of aripiprazole in children and adolescents with irritability associated with autistic disorder. J Am Acad Child Adolesc Psychiatry. 2009;48(11):1110-1119.
PubMed | Link to Article
4
Owen R, Sikich L, Marcus RN, et al. Aripiprazole in the treatment of irritability in children and adolescents with autistic disorder. Pediatrics. 2009;124(6):1533-1540.
PubMed | Link to Article
5
Aman MG, Kasper W, Manos G, et al. Line-item analysis of the Aberrant Behavior Checklist: results from two studies of aripiprazole in the treatment of irritability associated with autistic disorder. J Child Adolesc Psychopharmacol. 2010;20(5):415-422.
PubMed | Link to Article
6
Shea S, Turgay A, Carroll A, et al. Risperidone in the treatment of disruptive behavioral symptoms in children with autistic and other pervasive developmental disorders. Pediatrics. 2004;114(5):e634-e641.
PubMed | Link to Article
7
Martin A, Scahill L, Anderson GM, et al. Weight and leptin changes among risperidone-treated youths with autism: 6-month prospective data. Am J Psychiatry. 2004;161(6):1125-1127.
PubMed | Link to Article
8
Aman MG, Arnold LE, McDougle CJ, et al. Acute and long-term safety and tolerability of risperidone in children with autism. J Child Adolesc Psychopharmacol. 2005;15(6):869-884.
PubMed | Link to Article
9
Bobo WV, Cooper WO, Stein CM, et al. Antipsychotics and the risk of type 2 diabetes mellitus in children and youth. JAMA Psychiatry. 2013;70(10):1067-1075.
PubMed | Link to Article
10
Galling B, Roldán A, Nielsen RE, et al. Type 2 diabetes mellitus in youth exposed to antipsychotics: a systematic review and meta-analysis. JAMA Psychiatry. 2016;73(3):247-259.
PubMed | Link to Article
11
Lindsay RL, Leone S, Aman MG. Discontinuation of risperidone and reversibility of weight gain in children with disruptive behavior disorders. Clin Pediatr (Phila). 2004;43(5):437-444.
PubMed | Link to Article
12
Hellings JA, Zarcone JR, Crandall K, Wallace D, Schroeder SR. Weight gain in a controlled study of risperidone in children, adolescents and adults with mental retardation and autism. J Child Adolesc Psychopharmacol. 2001;11(3):229-238.
PubMed | Link to Article
13
Research Units on Pediatric Psychopharmacology Autism Network. Risperidone treatment of autistic disorder: longer-term benefits and blinded discontinuation after 6 months. Am J Psychiatry. 2005;162(7):1361-1369.
PubMed | Link to Article
14
Bouza C, López-Cuadrado T, Gutierrez-Torres LF, Amate J. Efficacy and safety of metformin for treatment of overweight and obesity in adolescents: an updated systematic review and meta-analysis. Obes Facts. 2012;5(5):753-765.
PubMed | Link to Article
15
Yanovski JA, Krakoff J, Salaita CG, et al. Effects of metformin on body weight and body composition in obese insulin-resistant children: a randomized clinical trial. Diabetes. 2011;60(2):477-485.
PubMed | Link to Article
16
Zheng W, Li XB, Tang YL, Xiang YQ, Wang CY, de Leon J. Metformin for weight gain and metabolic abnormalities associated with antipsychotic treatment: meta-analysis of randomized placebo-controlled trials. J Clin Psychopharmacol. 2015;35(5):499-509.
PubMed | Link to Article
17
Jarskog LF, Hamer RM, Catellier DJ, et al; METS Investigators. Metformin for weight loss and metabolic control in overweight outpatients with schizophrenia and schizoaffective disorder. Am J Psychiatry. 2013;170(9):1032-1040.
PubMed | Link to Article
18
Wu RR, Jin H, Gao K, et al. Metformin for treatment of antipsychotic-induced amenorrhea and weight gain in women with first-episode schizophrenia: a double-blind, randomized, placebo-controlled study. Am J Psychiatry. 2012;169(8):813-821.
PubMed | Link to Article
19
Klein DJ, Cottingham EM, Sorter M, Barton BA, Morrison JA. A randomized, double-blind, placebo-controlled trial of metformin treatment of weight gain associated with initiation of atypical antipsychotic therapy in children and adolescents. Am J Psychiatry. 2006;163(12):2072-2079.
PubMed | Link to Article
20
Shin L, Bregman H, Breeze JL, Noyes N, Frazier JA. Metformin for weight control in pediatric patients on atypical antipsychotic medication. J Child Adolesc Psychopharmacol. 2009;19(3):275-279.
PubMed | Link to Article
21
Morrison JA, Cottingham EM, Barton BA. Metformin for weight loss in pediatric patients taking psychotropic drugs. Am J Psychiatry. 2002;159(4):655-657.
PubMed | Link to Article
22
Arman S, Sadramely MR, Nadi M, Koleini N. A randomized, double-blind, placebo-controlled trial of metformin treatment for weight gain associated with initiation of risperidone in children and adolescents. Saudi Med J. 2008;29(8):1130-1134.
PubMed
23
Gonçalves P, Araújo JR, Martel F. Antipsychotics-induced metabolic alterations: focus on adipose tissue and molecular mechanisms. Eur Neuropsychopharmacol. 2015;25(1):1-16.
PubMed | Link to Article
24
Ibrahim SH, Voigt RG, Katusic SK, Weaver AL, Barbaresi WJ. Incidence of gastrointestinal symptoms in children with autism: a population-based study. Pediatrics. 2009;124(2):680-686.
PubMed | Link to Article
25
World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310(20):2191-2194.
PubMed | Link to Article
26
Lord C, Rutter M, DiLavore PC, Risi S. Autism Diagnostic Observation Schedule: A Standardized Observation of Communicative and Social Behavior. Los Angeles, CA: Western Psychological Services; 1999.
27
American Psychiatric Association. Diagnostic and Statistical Manual of Mental Health Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.
28
Roid GH. Stanford-Binet Intelligence Scales. 5th ed. Itasca, IL: Riverside Publishing; 2003.
29
Mullen EJ. Mullen Scales of Early Learning. Bloomington, MN: Pearson Assessments; 1989.
30
Aman MG, Singh NN, Stewart AW, Field CJ. The aberrant behavior checklist: a behavior rating scale for the assessment of treatment effects. Am J Ment Defic. 1985a;89(5):485-491.
PubMed
31
Aman MG, Pearson DA. Monitoring and measuring drug effects, II: behavioral, emotional, and cognitive effects. In: Werry JS, Aman MG. Practitioner’s Guide to Psychoactive Drugs for Children and Adolescents. 2nd ed. New York, NY: Plenum Press; 1999:99-164.
32
Greenhill LL, Vitiello B, Fisher P, et al. Comparison of increasingly detailed elicitation methods for the assessment of adverse events in pediatric psychopharmacology. J Am Acad Child Adolesc Psychiatry. 2004;43(12):1488-1496.
PubMed | Link to Article
33
Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412-419.
PubMed | Link to Article
34
Mazefsky CA, Schreiber DR, Olino TM, Minshew NJ. The association between emotional and behavioral problems and gastrointestinal symptoms among children with high-functioning autism. Autism. 2014;18(5):493-501.
PubMed | Link to Article
35
Nikolov RN, Bearss KE, Lettinga J, et al. Gastrointestinal symptoms in a sample of children with pervasive developmental disorders. J Autism Dev Disord. 2009;39(3):405-413.
PubMed | Link to Article
36
McGuire K, Fung LK, Hagopian L, et al. Irritability and problem behavior in autism spectrum disorder: a practice pathway for pediatric primary care. Pediatrics. 2016;137(suppl 2):S136-S148.
PubMed | Link to Article
37
Aman MG, Singh NN, Stewart AW, Field CJ. Psychometric characteristics of the aberrant behavior checklist. Am J Ment Defic. 1985b;89(5):492-502.
PubMed
38
Guy W. Dosage Record and Treatment Emergent Symptoms Scale, ECDEU Assessment Manual for Psychopharmacology (Revised). Washington, DC: US Dept of Health Education & Human Welfare; 1976:223-244.
39
Kieser M, Friede T. Simple procedures for blinded sample size adjustment that do not affect the type I error rate. Stat Med. 2003;22(23):3571-3581.
PubMed | Link to Article
40
Fitzmaurice G, Laird N, Ware J. Applied Longitudinal Analysis. 2nd ed. Hoboken, NJ: John Wiley & Sons; 2011.
41
Valeri L, Vanderweele TJ. Mediation analysis allowing for exposure-mediator interactions and causal interpretation: theoretical assumptions and implementation with SAS and SPSS macros. Psychol Methods. 2013;18(2):137-150.
PubMed | Link to Article
42
McElroy SL, Frye MA, Altshuler LL, et al. A 24-week, randomized, controlled trial of adjunctive sibutramine versus topiramate in the treatment of weight gain in overweight or obese patients with bipolar disorders. Bipolar Disord. 2007;9(4):426-434.
PubMed | Link to Article
43
Penzner JB, Dudas M, Saito E, et al. Lack of effect of stimulant combination with second-generation antipsychotics on weight gain, metabolic changes, prolactin levels, and sedation in youth with clinically relevant aggression or oppositionality. J Child Adolesc Psychopharmacol. 2009;19(5):563-573.
PubMed | Link to Article

JAMA Psychiatry. Published online August 24, 2016. doi:10.1001/jamapsychiatry.2016