Email Address

In the field below enter the email address where you received the invitation letter.

Set & Continue

MEDICATIONS & NUTRITION: How parents and doctors decide whether to use medications, and research on nutrition



Evidence-Based research on nutrition:

The Role of Nutrition in Treating ASD

Many theories have been proposed about the possible role of nutrition in some or all of the behavioral issues people with ASD experience. Some of these theories are based on physiological differences that have been observed between people with ASD and those considered neurotypical, but the physiological or biochemical basis for other theories is unclear. Nevertheless the theories have spawned a large number of studies on the effects of providing or eliminating particular nutrients on behaviors associated with ASD.

This section reviews the most recent scientific evidence about several of the more frequently studied dietary changes, namely, the use of gluten-free and/or casein-free diets, omega-3 fatty acid or probiotic supplements, as well as research on feeding difficulties.

Gluten-free/Casein-free Diets

Several theories have been proposed for why elimination of gluten (a protein found in wheat and other grain products) and casein (one of the two primary proteins in cow’s milk) might help with symptoms of ASD. Persons with ASD are likely to suffer from gastrointestinal (GI) disturbances (such as stomach discomfort and bloating), for which dairy and gluten-containing products have recently received blame. However, it’s not clear that GI disturbances are more prevalent among persons with ASD, and these gastrointestinal disturbances have not consistently been associated with the behavioral and cognitive issues that define ASD. It should also be noted that diets involving the elimination of gluten, casein, or both are extremely restrictive, difficult to follow, and can result in severe deficiencies of other nutrients if not properly managed with the help of an experienced nutritionist or registered dietitian. Most importantly, as the following studies have found, gluten- and/or casein-free diets have shown no evidence of helping with symptoms of ASD.

A 2014 systematic review (Mari-Bauset, 2014)1 found 21 studies published from 1970 to 2013 that assessed the effects of a gluten-free and/or casein-free diet on persons diagnosed with an ASD. Only four of the studies were judged to be of high quality, meaning that the study participants were randomly assigned to the experimental diet or to a control diet and that the participants, their caregivers, and researchers were blinded to the dietary assignment until after completion of the study, among other factors. Some other studies included only one person or assigned all participants to the same diet. The authors found that the higher quality studies showed no evidence for a positive effect of the diet on any behavior associated with ASD, and that only the poorest quality studies (usually those studies with no control group and enrolling as few as one participant) showed any evidence favoring eliminating gluten- and casein-containing foods.

Subsequent to publication of the 2014 systematic review, a group of Indonesian researchers carried out a study in which they randomized 74 children with ASD, high levels of what they termed “maladaptive behavior,” and high levels of a chemical in their intestines that might serve as an indicator of poor digestive function, to eat a snack throughout the day that was high in gluten and casein or to eat an identical looking and tasting, but gluten and casein-free snack.2 Neither the parents, the children, nor the researchers were aware of which children were assigned to which snacks. Although the children’s baseline diet (the other food the children ate) was not described, it can be assumed they were allowed to eat whatever they wished. A measure of “maladaptive behavior” was employed. After one week, maladaptive behavior decreased equally in both groups of children. Thus, the researchers concluded that gluten and casein play no role in the undesirable behaviors of children with ASD.

A well-designed US study published in 20163 showed that dietary challenges had no statistically significant effects on measures of physiologic functioning, behavior problems, or autism symptoms. This study restricted gluten and casein for 6 weeks from the diets of 14 children (3 to 5 years of age) with confirmed ASD diagnoses, then randomly divided the children into four groups and gave each group weekly snacks that contained gluten, casein, gluten and casein, or placebo (because it has been shown that negative reactions to foods nearly always subside within a week). This portion of the study continued for 12 weeks, after which the children were studied for an additional 12 weeks with dietary monitoring. In addition to employing a randomized controlled trial design, the researchers ensured that all participants were in a stable educational program using similar applied behavioral analysis (ABA) methodology, minimizing the potential influence of changes in other treatments at the same time as the dietary changes. Registered dietitians provided nutritional counseling to help families implement the diet and monitored the nutritional sufficiency of the children’s diets. Children’s adherence to the diet was measured repeatedly, as adherence to study diets is often poor and the effects of an intervention can only be assessed if adherence is excellent. The researchers also included both physiologic and behavioral outcome measures (both behaviors associated with ASD and behaviors that are not unique to ASD, e.g., sleep disturbance and overactivity). The only real limitation to this study was the small sample size, but combined with the findings of the systematic review and other recent study, the evidence is fairly unequivocal that this difficult, potentially risky diet is without benefit for children with ASD.

The Microbiome and the Use of Probiotics for ASD

“The microbiome” is the term used to describe the world of tiny organisms—single-celled bacteria and yeasts —that inhabit our intestines. We know little about their origin and less about their function, but what we do know is that the balance among the various species can be disrupted (by illness, medications changes in the foods we eat, and other stresses) and that these disruptions can result in GI disturbances, sometimes mild but occasionally severe. Based on the observation that people with ASD often seem to have GI disturbances, some researchers have theorized that perhaps the microbiomes of people with ASD differ innately from those of neurotypical people and if so, perhaps “normalizing” the microbiome might improve some of the behavioral or cognitive effects of ASD.

A number of studies have been done to assess whether the microbiome of people with ASD differs from that of people without ASD. A 2013 systematic review identified 15 such studies of small groups of people around the world. Eleven of these 15 studies (with a combined sample of 562 individuals) reported significant differences between ASD children and non-ASD children in the prevalence of various groups of GI bacteria in their intestines. Unfortunately the methods used to conduct each of the studies was so different that the authors could not pool any of the studies to conduct a meta-analysis (a mathematical procedure that allows the results of very similar studies to be combined, as if they were one large study) (Cao and colleagues, 2013).4 The meaning of the differences between ASD children and non-ASD children is entirely unclear. It could be a result of ASD, and not a cause: The genetic abnormalities that contribute to the behavioral, neurological, and cognitive manifestations of ASD could also cause imbalances in intestinal physiology that result in different microorganisms inhabiting the intestines. Or it could be the effect of particular medications children with ASD tend to take, or the effect of the limited diets on which some children with ASD subsist because of feeding difficulties. And the fact that only 11 of the 15 studies saw a difference makes the finding less interesting. But some researchers have made the leap to theorizing that maybe the effects of ASD could be improved by altering the microbiome of children with ASD using probiotics.

Probiotics is the term for the microorganisms that inhabit the microbiome and are thought to confer health on their host. They are the so-called good bacteria. Numerous studies have been conducted to assess whether these good bacteria can actually be used to treat various diseases or conditions. High quality studies are very difficult to conduct to answer these questions. Not surprisingly the results have varied widely, so as difficult as they are to conduct, they are even more difficult to interpret. A group of physicians, behavioral psychologists, and other experts in ASD have formed a consortium that met with parents of children with ASD in 2014 to examine the research that has been done so far and to define the components of a high-quality experiment ( ). In 2015, they published the guidelines they developed for conducting studies of probiotics in children with ASD.(Frye, 2015)5 NO rigorous scientific studies have yet examined the effects of giving probiotics to children with ASD: As of February 2016, only two clinical trials to study the effects of probiotics on children with ASD are registered with the federal government’s website,, and both trials are still recruiting prospective participants.

Omega-3 Fatty Acids and ASD

Omega-3 fatty acids are one of a group of molecules needed to synthesize the membranes that surround each of the millions of cells that make up our bodies. Because we cannot make the most basic of the omega-3 fatty acids, it is considered essential and we must get it from the foods we eat (flax seed, canola oil, walnuts, salmon and other dark fish, fish oil). Omega-3 fatty acids are considered to be important for the optimal development of the nervous system, and limited exposure to maternal omega-3 fatty acids have been implicated in the heightened risk for delayed cognitive development associated with preterm birth. No studies have shown an association between deprivation of omega-3 fatty acids and the development of ASD. Still, some researchers have theorized that supplementation of pregnant or nursing women or their infants will help prevent or treat ASD.

A 2011 systematic review tried to identify all high-quality studies of the effects of using omega-3 fatty acids to treat children with ASD.(James, 2011)6 They found only 2 studies (with a total of 37 children) that met their criteria of including a control (placebo)group, randomly assigning children to the treatment and control groups, and ensuring that parents, caregivers, and testers were blinded to the treatment! Six other studies were excluded because they failed to meet one or more of these criteria. Neither of the 2 studies found an effect of omega-3 fatty acid supplements on tests for social interaction, communication, stereotypic behaviors (like hand flapping), or hyperactivity. Since that review was published, a small number of additional good-quality studies have appeared.

Voigt and his colleagues conducted a 6-month study to determine whether supplementing children diagnosed with ASD with one particular omega-3 fatty acid, docosahexaenoic acid (DHA) improves behavior.(Voigt, 2014)7 They randomized 48 children, 3 to 10 years of age to receive a low dose of DHA (0.2 grams) daily. Behavior was assessed at the beginning of the study and at the end of 3 and 6 months using standard measures (the Clinical Global Impressions-Improvement scale to the Child Development Inventory and the Aberrant Behavior Checklist, and the Behavior Assessment Scale for Children (BASC)) by parents and testers who were blinded to treatment. The researchers found no effect of DHA on the children’s core symptoms.

A Canadian study published in 2015 randomized 38 children 2 to 5 years old and diagnosed with ASD to 1.5 grams per day of a commercial brand fish-oil supplement for 6 months. The children’s symptom severity was assessed at the beginning of the study (baseline) and periodically using the Pervasive Developmental Disorder Behavioral Inventory (PDDBI). The effect of omega-3 fatty acids on externalizing behaviors was measured using the Behavior Assessment System for Children, Second Edition (BASC-2). Additional outcome measures included the Vineland Adaptive Behavior Scales, Second Edition (VABS-II) at baseline, week 12, and week 24; the Preschool Language Scale-4 (PLS-4) at baseline and week 24; the Clinical Global Impression-Improvement (CGI-I) scale, and a measure of omega-3 fatty acids in the blood at each visit to assess global improvement. Blood levels of omega-3 fatty acids increased in the children who received the supplement compared with the control children. Nevertheless, the authors reported no significant difference between groups on the 0- to 24-week change in PDDBI autism composite scores, and the BASC-2 showed a worsening in scores over time in the group treated with the fish oil but not the control group. No effect was seen over time in either adaptive function or language.(Mankad, 2015) 8

Although neither of these trials seems to have monitored other components of the children’s diets or ongoing behavioral treatment over the six months of the interventions, differences in diet and treatment should have been balanced by the randomization. Thus the results of the review as well as these two more recent studies suggest that supplementation with omega-3 fatty acids is not effective in treating ASD.

Feeding Difficulties or Food Selectivity

Feeding difficulties or food selectivity are terms used to refer to behaviors ranging from food refusal to a limited set of food preferences. The observation that food selectivity seems to be prevalent among children with ASD fits with an overall pattern of restricted, repetitive behaviors among those with ASD, but are feeding difficulties and food selectivity more prevalent among those with ASD than among the general pediatric population? And what methods, if any, help to treat the condition, which left untreated, could potentially result in nutrient deficiencies?

Mari Bauset and colleagues conducted a systematic review of the literature to assess the prevalence of food selectivity among children with ASD and the types of treatments.9 They identified 25 studies published between 1970 and 2013 on feeding difficulties in children with ASD. The reported prevalence in the larger studies was greater than 50%, compared with about 25% in the general population, supporting the idea that food selectivity is more prevalent among this group of children. A small number of studies reported on treatment methods and showed support for behavioral methods.

Marshall and colleagues also systematically reviewed methods to address feeding difficulties in children with ASD.10 They identified 23 studies conducted between 2000 and 2013. All employed what are called operant conditioning techniques (these are techniques that progressively shape desired behaviors by using positive reinforcement). The studies generally succeeded in increasing the volume of food eaten but not the variety. But, because none of the studies enrolled more than 5 children and none employed good study design (that is, a randomized controlled trial), the evidence they present is somewhat weak.

Sharp and colleagues developed and evaluated a behaviorally based parent training program intervention to treat feeding difficulties in children with ASD called the Autism MEAL Plan.11 The intervention involved 8 1-hour group educational sessions that covered general behavioral principles (e.g., routine and consistency), methods of creating a pleasant mealtime setting, behavioral methods for introducing new foods and extinguishing aversive reactions to familiar foods, and methods to promote self feeding. Homework was assigned each week to encourage parents to apply and practice the new skills with their children and these experiences were reviewed in the following session. To assess the program, the authors conducted a pilot randomized controlled trial with 19 families (10 participated in the intervention and 9 on the waiting list for the program served as the controls). The program had high acceptance, improved some negative feeding behaviors, was perceived by parents to be effective, and reduced caregiver levels of stress.

Marshall and colleagues conducted a randomized controlled trial to compare the effectiveness of two different interventions to improve increase dietary variety and intake and reduce other difficult mealtime behaviors among 68 children 2 to 6 years old with ASD or with what they called non-medically complex histories. The participants were randomized to one of two types of treatment: One was based on operant conditioning (shaping desired behaviors by rewards) and the other on systematic desensitization (presenting an aversive food repeatedly, with modeling and play, until the reaction is extinguished). Each intervention was 10 sessions, as frequently as the parents desired. At the end of 3 months, both groups of children showed significant improvements, although those in the operant conditioning group showed greater increases in the variety of foods eaten than those in the desensitization group.(Marshall et al, 2015)12

These studies provide evidence that programs based on behavioral techniques can help decrease feeding difficulties in children with ASD. The programs require strong motivation and commitment from families, and the content of current programs will need to undergo further modification and testing.

Use of Digestive Enzymes and Hormones

Based on an unproven theory that the cause of ASD might lie in poor digestion of certain foods, purportedly resulting in the abnormal buildup of carbohydrate breakdown products with opium-like effects, several bodies of research have tried to determine whether supplementing children with ASD with digestive enzymes or secretin, a GI hormone involved in digestion decreases behaviors associated with ASD.

Two relatively recent randomized controlled trials tested the effects of giving children with ASD digestive enzymes. A double-blind placebo controlled trial conducted in Australia and published in 2010 randomized 43 children to receive a commercial brand of prescription grade digestive enzymes or a placebo for 3 months; at the end of that time, children who were receiving the enzymes were switched to the placebo and vice versa, with blinding maintained. The study found no effect of the enzymes on any measure of children’s behavior with the exception of a small effect on the variety of foods eaten.13 A 2015 placebo-controlled trial conducted in Egypt randomized 101 children to receive a digestive enzyme preparation or a placebo for 3 months.14 The enzymes were not identified by name or contents. This study reported small but significant improvements in emotional response, general impression autistic score, general behavior, and gastrointestinal symptoms in the children who received the enzymes, but the lack of identification of the intervention product and several other aspects of the study are cause for concern. High quality prescription grade digestive enzymes are very costly and should be administered only under the supervision of a physician, preferably as part of a clinical trial. Only one trial of the effects of digestive enzymes on children with ASD has been registered with, the enzymes were administered with many other supplements, and the results have not been published.

Secretin is a hormone that is produced in the GI tract and aids digestion. Based on an incidental observation that several children with ASD who received intravenous secretin prior to undergoing a GI biopsy procedure exhibited some short-term improvement in their ASD behaviors, numerous studies have been conducted on the effects of IV secretin on children with ASD (the hormone must be injected because it is a small protein and would be digested if given orally). In 2012, Williams and colleagues conducted a rigorous systematic review of 14 trials of secretin for ASD (enrolling over 900 children in total) that met the inclusion criteria (randomized placebo-controlled trials that used validated tests for ASD).15 Their analysis found no positive effects of secretin administration on behaviors associated with ASD. Based on their findings, they recommended that no further studies of secretin be conducted on children with ASD unless a type of ASD could be identified whose origin proved to involve the hormone.

1  Mari-Bauset S, Zazpe I, Mari-Sanchis A, et al. Evidence of the gluten-free and casein-free diet in autism spectrum disorders: a systematic review.  Child Neurol. 2014 Dec;29(12):1718-27. PMID: 24789114. (subscription needed)

2  Pusponegoro HD, Ismael S, Firmansyah A, et al. Gluten and casein supplementation does not increase symptoms in children with autism spectrum disorder.  Acta Paediatr. 2015 Jul 4PMID: 26148018. (subscription needed)

3  Hyman SL, Stewart PA, Foley J, et al. The Gluten-Free/Casein-Free Diet: A Double-Blind Challenge Trial in Children with Autism.  J Autism Dev Disord. 2015 Sep 5PMID: 26343026. (subscription needed)

4  Cao X, Lin P, Jiang P, et al. Characteristics of the gastrointestinal microbiome in children with autism spectrum disorder: a systematic review.  Shanghai Arch Psychiatry. 2013 Dec;25(6):342-53. PMID: 24991177. (subscription needed)

5  Frye RE, Slattery J, MacFabe DF, et al.  Approaches to studying and manipulating the enteric microbiome to improve autism symptoms. (download pdf)  Microb Ecol Health Dis. 2015;26:26878. PMID: 25956237.

6 James S, Montgomery P, Williams K. Omega-3 fatty acids supplementation for autism spectrum disorders (ASD).  Cochrane Database Syst Rev. 2011(11):Cd007992. PMID: 22071839. (subscription needed)

7  Voigt RG, Mellon MW, Katusic SK, et al. Dietary docosahexaenoic acid supplementation in children with autism.  J Pediatr Gastroenterol Nutr. 2014 Jun;58(6):715-22. PMID: 24345834. (subscription needed)

8  Mankad D, Dupuis A, Smile S, et al.  A randomized, placebo controlled trial of omega-3 fatty acids in the treatment of young children with autism. (download pdf)  Mol Autism. 2015;6:18. PMID: 25798215.

9  Mari-Bauset S, Zazpe I, Mari-Sanchis A, et al. Food selectivity in autism spectrum disorders: a systematic review.  J Child Neurol. 2014 Nov;29(11):1554-61. PMID: 24097852. (subscription needed)

10  Marshall J, Ware R, Ziviani J, et al. Efficacy of interventions to improve feeding difficulties in children with autism spectrum disorders: a systematic review and meta-analysis.  Child Care Health Dev. 2015 Mar;41(2):278-302. PMID: 24962184. (subscription needed)

11  Sharp WG, Burrell TL, Jaquess DL. The Autism MEAL Plan: a parent-training curriculum to manage eating aversions and low intake among children with autism.  Autism. 2014 Aug;18(6):712-22. PMID: 24101716. (subscription needed)

12  Marshall J, Hill RJ, Ware RS, et al.Multidisciplinary intervention for childhood feeding difficulties.  J Pediatr Gastroenterol Nutr. 2015 May;60(5):680-7. PMID: 25534777.(subscription needed)

13  Munasinghe SA, Oliff C, Finn J, et al. Digestive enzyme supplementation for autism spectrum disorders: a double-blind randomized controlled trial.  J Autism Dev Disord. 2010 Sep;40(9):1131-8. PMID: 20204691. (subscription needed)

14  Saad K, Eltayeb AA, Mohamad IL, et al.  A Randomized, Placebo-controlled Trial of Digestive Enzymes in Children with Autism Spectrum Disorders. (download pdf)  Clin Psychopharmacol Neurosci. 2015 Aug 31;13(2):188-93. PMID: 26243847.

15  Williams K, Wray JA, Wheeler DM. Intravenous secretin for autism spectrum disorders (ASD).  Cochrane Database Syst Rev. 2012;4:Cd003495. PMID: 22513913. (subscription needed)

Evidence-Based research on medications:

  • Oxytocin and autism: a systematic review of randomized controlled trials 16 (subscription needed)
    This article systematically reviews all clinical trials of the medication oxytocin in persons with Autism Spectrum Disorders (ASD) as of March, 2013. Oxytocin is a naturally occurring substance synthesized in the human hypothalamus and released into the bloodstream by the pituitary gland. In synthetic form, it is often used for inducing labor or managing postpartum hemorrhage. Because of some positive results in non-autistic individuals, oxytocin was hypothesized to improve social symptoms of ASD. Neither this drug nor any others are FDA approved for this use.

    Six randomized controlled trials were identified. A total of 101 persons with ASD were enrolled; only six females were included. All trials lasted six weeks or less. Some studies utilized intranasal administration while others administered oxytocin intravenously; doses varied by study. Studies used different outcomes to measure effectiveness. The studies’ risk of bias, based on the randomization process, blinding, and attrition rate, ranged from medium to high. Although some studies reported potentially promising findings, results were inconsistent. No adverse events requiring medical attention were reported. Additional research with larger population samples and more females must be conducted before this treatment can be recommended.

  • Mitigation of sociocommunicational deficits of autism through oxytocin-induced recovery of medial prefrontal activity: A randomized trial. 17
    Researchers in Japan conducted a randomized clinical trial (RCT) of oxytocin to assess its efficacy in mitigating social and communication deficits in high functioning men with autism spectrum disorders (ASD). They used a within-subject crossover design, where 40 men were randomized to receive a single dose of intra-nasal oxytocin or placebo before a psychological task.

    The researchers created a psychological task which required participants to make judgments regarding the intentions of others based on communicational content in which verbal and nonverbal information conflicted. The men viewed short videos in which professional actors spoke an emotional word (verbal information) with an emotional facial expression and expressive voice (nonverbal information). The videos consisted of two types of emotionally congruent videos with negative (NV−V−) or positive (NV+V+) nonverbal and verbal information and two types of incongruent videos with negative nonverbal and positive verbal (NV−V+) or positive nonverbal and negative verbal (NV+V−) information. On the basis of the video content participants were asked to make a “friend or foe” judgment of the actor. Persons with ASD tend to respond less to nonverbal communication; the hypothesis was that frequency of nonverbal information–based judgments (NVJs), the response time for NVJs, and brain activity of the medial prefrontal cortex (measured by Functional Magnetic Resonance Imaging) during NVJs would be higher with oxytocin than placebo.

    Compared to placebo, intranasal oxytocin led to significantly more frequent NVJs, with significantly shorter response time. The magnitude of neural effects was predictive of the behavioral effects. The authors concluded that intranasal administration of oxytocin enables highly functioning individuals with ASD to exhibit more typical and smoother behavioral responses to social communication when verbal and nonverbal information conflicts.

  • 16 Preti A, Melis M, Siddi S, Vellante M, Doneddu G, Fadda R. Oxytocin and autism: a systematic review of randomized controlled trials. Journal of child and adolescent psychopharmacology.  Mar 2014;24(2):54-68. Journal of autism and developmental disorders. Jan 2013;43(1):57-67. (subscription needed)

    17 Watanabe T, Abe O, Kuwabara H, Yahata N, Takano Y, Iwashiro N, et al. Mitigation of sociocommunicational deficits of autism through oxytocin-induced recovery of medial prefrontal activity: A randomized trial.  JAMA Psychiatry S2- A M A Archives of General Psychiatry S2- Archives of General Psychiatry. 2014;71(2):166-75. (subscription needed)

General Guides:

Hide Comments