See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/262024982

Evidence of the Gluten-Free and Casein-Free Diet in Autism Spectrum

Disorders: A Systematic Review

Article in Journal of Child Neurology · April 2014

DOI: 10.1177/0883073814531330 · Source: PubMed

CITATIONS READS
118 3,596

5 authors, including:

Salvador Marí-Bauset Itziar Zazpe

University of Valencia Universidad de Navarra
22 PUBLICATIONS 495 CITATIONS 95 PUBLICATIONS 3,334 CITATIONS

SEE PROFILE SEE PROFILE

Amelia Mari-Sanchis Maria Morales Suárez-Varela

Complejo Hospitalario de Navarra University of Valencia
24 PUBLICATIONS 641 CITATIONS 378 PUBLICATIONS 5,241 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Public Health Reviews View project

SENDO project View project

All content following this page was uploaded by Salvador Marí-Bauset on 28 July 2015.

The user has requested enhancement of the downloaded file.

 Journal http://jcn.sagepub.com/
of Child Neurology

Evidence of the Gluten-Free and Casein-Free Diet in Autism Spectrum Disorders: A Systematic
Review
Salvador Marí-Bauset, Itziar Zazpe, Amelia Mari-Sanchis, Agustín Llopis-González and María Morales-Suárez-Varela
J Child Neurol published online 30 April 2014
DOI: 10.1177/0883073814531330

The online version of this article can be found at:

http://jcn.sagepub.com/content/early/2014/04/29/0883073814531330

Published by:

http://www.sagepublications.com

Additional services and information for Journal of Child Neurology can be found at:

Email Alerts: http://jcn.sagepub.com/cgi/alerts

Subscriptions: http://jcn.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - Apr 30, 2014

What is This?

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

Topical Review Article
Journal of Child Neurology
1-10
Evidence of the Gluten-Free and Casein- ª The Author(s) 2014
Reprints and permission:
sagepub.com/journalsPermissions.nav
Free Diet in Autism Spectrum Disorders: DOI: 10.1177/0883073814531330
jcn.sagepub.com
A Systematic Review

Salvador Marı́-Bauset, DPharm1, Itziar Zazpe, MD, PhD2,3,

Amelia Mari-Sanchis, MD4, Agustı́n Llopis-González, PhD1,5,
and Marı́a Morales-Suárez-Varela, MD, PhD1,5

Abstract
In autism spectrum disorders, many parents resort to alternative treatments and these are generally perceived as risk free. Among
these, the most commonly used is the gluten-free, casein-free diet. The objective of this work was to conduct a systematic review
of studies published from 1970 to date related to the gluten-free, casein-free diet in autism spectrum disorder patients. Few stud-
ies can be regarded as providing sound scientific evidence since they were blinded randomized controlled trials, and even these
were based on small sample sizes, reducing their validity. We observed that the evidence on this topic is currently limited and
weak. We recommend that it should be only used after the diagnosis of an intolerance or allergy to foods containing the allergens
excluded in gluten-free, casein-free diets. Future research should be based on this type of design, but with larger sample sizes.

Keywords
gluten-free, casein-free diet, autism, autism spectrum disorders (ASDs), review

Received January 13, 2014. Accepted for publication March 18, 2014.

Autism spectrum disorders are a clinically characterized by dif- breast milk, yogurt, cheese, butter, cream or ice cream, among
ficulties with reciprocal social interactions; verbal and nonver- others.
bal communication deficiencies; and restricted, repetitive, and On the other hand, in relation to children with autism spec-
stereotyped behaviors and interests.1 According to a recent trum disorders, these diets involve significant changes to their
publication analyzing 2008 data,2 the prevalence has increased routine and such changes can, in themselves, affect their eating
to 11.3 per 1000 people, and it is notably more common in men behaviors.9-11 Additionally, the adoption of elimination diets
(ratio 4:1). On the other hand, no significant differences have works against efforts to improve the social integration of such
been reported as a function of socioeconomic level or cultures.3 children, in that a personal diet is an isolating factor.12

Gluten-Free, Casein-Free Diet: Background

1
Unit of Public Health and Environmental Care, Department of Preventive
So far, there are no curative treatments for this disorder; how-
Medicine, University of Valencia, Valencia, Spain
ever, there is some hope of advances.4 In this context, many 2
Department of Nutrition and Food Sciences and Physiology, University of
parents have turned to alternative treatments5-7 driven by the Navarra, Navarra, Spain
3
frustration and concern caused by the diagnosis rather than with Biomedical Research Centre Network on Obesity and Nutrition (CIBERobn),
sound justification, and this issue is compounded by the fact Physiopathology of Obesity and Nutrition, Institute of Health Carlos III,
that the treatments are generally thought to be risk-free. These Madrid, Spain
4
Clinical Nutrition and Dietetics Unit, Navarra Hospital Complex, Navarra,
alternative treatments include the adoption of elimination diets, Spain
in particular the gluten-free, casein-free diet,8 the focus of this 5
CIBER Epidemiologia y Salud Pública (CIBERESP), Institute of Health Carlos
paper. III, Madrid, Spain
In relation to this, the elimination of gluten implies the
exclusion of all food items containing wheat, oats, barley or Corresponding Author:
Maria Morales-Suárez-Varela, MD, PhD, Unit of Public Health and
rye, that is, all flours, bread, rusks, pasta, pastries, and other Environmental Care, Department of Preventive Medicine. University of
bakery products made with these cereals, while the elimination Valencia, Avd. Vicente Andres Estellés s/n, Burjassot 46100, Valencia, Spain.
of casein means no intake of dairy products: milk, including Email: [email protected]

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

2 Journal of Child Neurology

Opioid Theory for Autism Spectrum Disorders groups, and Whiteley et al33 reported the appearance of a pos-
sible diet-related autism phenotype that seems to be emerging
The most commonly cited theory to justify adoption of a
supportive of a positive dietary effect with slight improvement
gluten-free, casein-free diet is related to neurotransmitters13
in certain groups with autism spectrum disorders. On the other
and concerns the release of peptides with an opioid activity
hand, Sponheim12 did not observe any improvement after intro-
in the intestines. After digestion, certain types of proteins could
duction of the elimination diet, but rather behavioral regression
cross the intestinal mucosa intact,14 if this were more perme-
due to stigmatization. Elder et al34 and Seung et al35 did not
able than normal—this being the case when it is impaired by
find any improvement in the behavior of participants in the
immunologic factors or by lesions in the case of celiac disease.
intervention group.
If these peptides, transported by the bloodstream, were to cross
Having discussed the questionable effectiveness of this
the blood-brain barrier and reach the central nervous system in
nutritional intervention on cognitive-behavioral function, we will
large quantities, it would affect brain functioning.15 The hydro-
now assess its safety. Cornish11 did not find any significant
lysis of proteins from cereals and milk would generate exogen-
nutritional differences between children with autism spectrum
ous neuropeptides (exorphines) such as gluteomorphins from
disorders as a function of whether they were on the gluten-free,
gluten and beta-casomorphins from casein.
casein-free diet, similar to the findings of Johnson et al.36 On the
It should, however, be highlighted that exorphins have a low
other hand, Arnold et al37 observed a significantly lower concen-
affinity for opioid receptors and that in dietary proteins there
tration of amino acids, including tryptophan in children with aut-
are also amino acid sequences with antagonist activity on
ism spectrum disorders on gluten-free, casein-free diets. Higher
opioid receptors which, despite having been known of for many
homocysteine levels have been observed in patients on a
years, tend not to be considered in this context.16
gluten-free diet long-term compared to typically developing chil-
What is more, experiments have failed to find abnormally
dren, and this implies deficiencies in folates and vitamin B6,
high concentrations of opioid peptides in either plasma or the
increasing cardiovascular risk in the medium and long term.38-40
nervous system of patients with autism spectrum disorders.17
Mariani et al41 reported that patients on a gluten-free diet had high
As for urinary excretion, urinary opioid peptides have not been
intakes of proteins and lipids but low intakes of carbohydrates,
detected in people with autism spectrum disorders using mod-
fiber, calcium, and iron. In line with this, Marcason42,43 warns
ern methods with great sensitivity and specificity (namely,
about the risk of gluten-free diets resulting in deficient intake of
mass spectrometry coupled with high-performance liquid
both macro- and micronutrients, the associated restrictions mak-
chromatography).18-20
ing it much more difficult to achieve a balanced diet than when a
broader variety of foods are consumed.
Prevalence Similarly, a casein-free diet could result in calcium defi-
ciency.44-47 Aldamiz-Echevarria et al48 indicated that 76% of
The adoption of gluten-free, casein-free diet, as an alternative patients on a casein-free diet had a total lipid intake within the
treatment, is a poorly studied phenomenon. In the literature, recommended range, but 85% had high ratios of o6/o3 and low
figures are highly variable, indicating that this approach is tried plasma levels of docosahexaenoic acid (DHA) and eicosapen-
in 20% to 70% of cases. For instance, Harrington et al6 reported taenoic acid (EPA), a similar pattern being described by Schu-
rates of 66%; Wong et al,21 30%; Herndon et al,22 31.1%; chardt et al.49 Further, slower bone development was observed
Bandini et al,23 20.7%; and Hall et al24 and Sharp et al,25 30%. in children with autism spectrum disorders on a casein-free
diet than among those without dietary restrictions,50 while
Neumayer et al51 demonstrated that children with autism spec-
Behavior and Physiological Perspective trum disorders had a lower bone density than controls. In this
The first author to establish an association between the fre- latter study, the total energy and macronutrient intakes did not
quency and severity of schizophrenia and the intake of foods differ significantly between groups, but the intakes of vitamin
containing gluten and dairy products was Dohan: the foods’ D and calcium were lower in children with autism spectrum
withdrawal improved symptoms and their reintroduction disorders and this can be attributable to lower consumption
worsened them.26,27 Subsequently, Panksepp28 suggested that or even the elimination of milk and other dairy products.
the behavioral changes associated with autism were the result All the above justifies this systematic review of the studies
of an abnormal activation of the opioid system because of an published since 1970 concerning dietary restriction and its
excess of agonists in the brain. It has been considered that glu- impact on autism spectrum disorders. Specifically, the
ten from cereal and casein from dairy products could be respon- objectives of this study were to determine, on the basis of the
sible, as they are a source of ‘‘exorphins,’’ peptides with opioid available scientific data, (1) the apparent efficacy and (2) any
activity.29-32 possible associated metabolic risks of dietary restrictions.
Considering publications since 1970, excluding theses or
book chapters, we found relatively few original studies on
elimination diets that analyze the impact of foods on behavior Methods
in autism spectrum disorders. Several of these studies demon- We conducted a systematic review of the medical literature related to
strated significant improvements in intervention vs control gluten-free, casein-free type diets. The date of the last search was

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

Marı´-Bauset et al 3

Table 1. Levels of Evidence.

Levels of
evidence Characteristics

1
1þþ High-quality meta-analyses, systematic reviews of RCTs, or RCTs with a very low risk of bias
1þ Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias
1– Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias
2
2þþ High-quality systematic reviews of case-control or cohort or studies.
2þ High-quality case-control or cohort studies with a very low risk of Confounding, bias, or chance and a high probability that the
relationship is casual
2– Well-conducted case-control or cohort studies with a low risk of confounding, bias, or chance and a moderate probability that
the relationship is casual
3 Nonanalytic studies, eg, case reports, case series
4 Expert opinion
Abbreviation: RCT, randomized controlled trial.
Source: SIGN, Scottish Intercollegiate Guidelines Network (2008).

Table 2. Grades of Recommendation.

Grades of
recommendation Definition

A At least 1 meta-analysis, systematic review, or RCT rated as 1þþ, and directly applicable to the target population; or a
systematic review of RCTs or a body of evidence consisting principally of studies rated as 1þ, directly applicable to
the target population, and demonstrating overall consistency of results
B A body of evidence including studies rated as 2þþ, directly applicable to the target population, and demonstrating
overall consistency of results; or extrapolated evidence from studies rated as 1þþ or 1þ

C A body of evidence including studies rated as 2þ, directly applicable to the target population, and demonstrating overall
consistency of results; or extrapolated evidence from studies rated as 2þþ

D Evidence level 3 or 4; or extrapolated evidence from studies rated as 2þ

Abbreviation: RCT, randomized and controlled trial.
Source: SIGN, Scottish Intercollegiate Guidelines Network (2008).

September 30, 2013. We based our search on the Medline database, in planning of dietary guidelines for patients with autism spectrum disor-
accordance with the proposals of the Spanish National Health System. ders. For this classification of the evidence and recommendations, we
In addition, we also consulted other databases: Cochrane Library, employed an instrument proposed by the Scottish Intercollegiate Guide-
Scielo, ScienceDirect, and Embase. For the searches, we used the lines Network.55 The scale proposes that 2 characteristics of the source be
keywords gluten-free, casein-free diet, autism, Autism Spectrum used for assessing the quality of the scientific evidence provided (level of
Disorders (ASDs), and review, with the corresponding Boolean opera- evidence): the study design and the risk of bias. Numbers from 1 to 4 are
tors. This paper complies with the methodological norms established used to rate the study design, whereas signs (þþ, þ, and –) indicate the
for the publication of systematic reviews52,53 and the PRISMA assessed risk of bias, according to the degree of fulfillment of key criteria
recommendations.54 related to this potential risk (Table 1). Based on this assessment of the
We first retrieved systematic reviews and full original articles pub- quality of the scientific evidence in the source, grades are used (Table 2)
lished from 1970 to 2013. These publications were then included in the to classify the strength of associated recommendations (A, B, C, and D).
analysis provided that the participants, of any age, met the Diagnostic In addition to the aforementioned system of levels, we considered
and Statistical Manual of Mental Disorders, Fourth Edition, Text Revi- the following features, as applicable, to assess the level of evidence
sion, criteria for autism spectrum disorders; that they were put on a diet provided by the selected articles: (A) Degree of homogeneity of the
excluding gluten, casein or both; and that the outcome variables were group studied (as determined by definitions and criteria applied);
related to the potential biomedical or behavioral symptoms of autism (B) use of a control group and the appropriateness of the selection;
spectrum disorders. We did not restrict the searches by language. On the (C) type of experimental design (randomized or not); (D) knowledge
other hand, studies in which the diet was not under supervision of of the intervention by patients, relatives, and other observers (open,
the researchers and any that did not report on health outcomes were simple or double blind trial); (E) nature of the dietary regimen (level
excluded. of strictness) and degree of adherence; (F) selection of assessment
To guide the evaluation of the data in the papers retrieved, we defined criteria, including the instruments used (questionnaires, scales, etc.),
levels of evidence, on the basis of their methodological quality in terms of for assessing changes in patient status under the treatment; and (G) the
the study design. We then established grades of recommendations for the presence of confounding factors including any types of pharmacologic

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

 4
Table 3. Studies Considering the Effectiveness of the Gluten-Free, Casein-Free Diet.

Author N Design Assessment criteria Results LEa GR

62
Goodwin et al, 1971 Cases: 15 children with ASDs Cross-over intervention with Blood biochemistry; EEG; Correlation between autism and 3 D
Age: 6-13 y; controls: 14; siblings: gluten load test after GF diet; behavioral malabsorption, and between
1-13 y randomized; open-label; gluten sensitivity and cognitive
experimental impairment
Bird et al, 197763 9-y-old child with ASD; no controls Case report; GFCF diet; Behavioral Diet does not affect behavior 3 D
experimental
ÓBannion et al, 197864 1 child with ASD; no controls Descriptive; case report Behavioral; dietary restriction Wheat, maize, tomatoes, sugar, 3 D
mushrooms, and dairy
products cause behavioral
problems in the child
McCarthy et al, 197965 8 patients with ASD; unknown age; Open label; gluten load after GF Clinical picture and intestinal No abnormal histologic findings 2–
no healthy controls diet biopsy in the intestine; no association
between autism and celiac
disease
Reichelt et al, 199066 15 patients with ASDs; 3-17 y old Open label; urinary peptide Peptiduria; Parent and caregivers Behavioral improvement after 3 D
levels; GFCF/GF/CF; cohort questionnaire with the treatment in more than
study retrospective comparison 50% of children
Knivsberg et al, 199029 15 patients with ASD; age 6-14 y; no Open label; urinary peptide Parent and caregiver reports; Early improvement (first 6 mo), 3 D
healthy control group levels; GF/CF; cohort study standardized assessment less evident after 1 y
questionnaires
Sponheim, 199112 4 adults and 3 children with ASDs; 6- Experimental; randomized; Parent and caregiver reports; No behavioral improvement 2þþ B
month GF diet double blind (gluten-free or standardized assessment
placebo 1 y); cohorts questionnaires
Lucarelli et al, 199567 36 patients with ASDs, 8-13 y old; Open label before and after Standardized assessment After the dietary restriction, 2–

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

controls: 20 children without dietary restriction; casein load questionnaire; measurement decrease in the specific Ig
ASDs for the casein load test; test: double blind, placebo- of Ig levels levels; behavioral
2-month CF diet controlled cohort study improvement; casein load test
results not conclusive
Knivsberg et al, 199568 Continuation of the authors’ Experimental; open label; urinary Peptiduria; parent and caregiver No clear improvement after 1 y 3 D
previous study; cf. Knivsberg et al, peptide levels; GF/CF; cohort reports; standardized
1990 study assessment questionnaires
Adams et al, 199769 1 boy and 1 girl with ASDs; both 3 y Series of cases: (1) very high Parents report Anecdotal improvement of the 3 D
old doses of Mg þ vitamin B6; general behavior
(2) GFCF for 2 y
Knivsberg et al, 199970 Girl with an ASD; 7 y old Case report; GF for 2 y; Parents report; standardized Behavioral improvement in 3 D
experimental. assessment questionnaires communication and socializing
(continued)
 Table 3. (continued)

Author N Design Assessment criteria Results LEa GR

9
Whiteley et al, 1999 Cases: 5 children with ASDs and 9 GF diet for 5 mo; experimental; Peptiduria; parents report; Behavioral improvement 2–
children with autism; no cohort study standardized assessment
randomized controls: 6 children, questionnaires
all diagnosed with autism, formed
a control group of children not
involved with any dietary
intervention; GF > 6 mo
Cade et al, 200010 150 children with ASDs aged Open label; gluten load test after Behavioral; peptiduria Improvement of the autistic signs 2–
between 3.5 and 16 y; no healthy GF diet; experimental and symptoms
controls
Knivsberg et al, 200271 Cases: 10 children with ASDs, GFCF Randomized; simple blind; GFCF Parents and caregiver report; Behavioral improvement in 2þ C
diet for a y; controls: 10 children diet for 1 y; experimental standardized assessment communication and socializing
with ASDs with normal diet; both questionnaires
groups with a mean age of 7.5 y
Elder et al, 200634 15 people with ASDs; aged 2-16 y Randomized, retrospective, Parents report. Standardized No statistically significant 2þþ B
old; GFCF diet for 12 wk; no double blind, cross-over assessment questionnaires; differences were observed
healthy controls (RCT), 6 wk under normal and urinary peptide levels

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

GFCF diet, alternately

Irvin 200672 1 boy with an ASD, 12 y old Case report, GFCF for 4 d Direct observation of the level of No behavioral change 3 D
aggression and destructive
behavior

5
 6
Table 3. (continued)

Author N Design Assessment criteria Result LE GR

Patel and Curtis, 10 children with ASDs and Open label, Experimental. Behavioral assessment by physicians/ Lower urinary concentration of 3 D
200773 ADHD aged between 4 and 10 Children received an integrated parents teachers; urinary heavy heavy metals; parents report a
y old; no healthy controls treatment based on nutritional metals. behavioral improvement
(GFCF) and environmental
changes, plus the chelating agents,
for 3-6 months, as well as the
usual behavioral therapy and
physiotherapy
Seung et al, 200735 13 children with ASDs; 2-16 y Retrospective, Video recording; assessment of No statistically significant differences 2þþ B
old; no healthy controls randomized, double blind, cross verbal and nonverbal were observed
over study; 6 wk on a normal and communication
GFCF diet, alternatively
Millward et al, 200859 Systematic review There is no empirical evidence base 1þ A
for recommending the GFCF diet
Hyman et al, 201056 30 children with ASDs; Experimental; double blind; cross- Behavioral assessment No significant differences in the 1þ A
30-45 mo over; RCT; 18 wk under GFCF preliminary results (2010); no
and normal diet alternately empirical evidence to recommend
the GFCF diet
Mulloy et al, 201060 Systematic review No empirical evidence to 1þ A
recommend the GFCF diet
Whiteley et al, 201074 72 children with ASDs; no Single-blind, randomized, clinical Standardized assessment Significant improvement in some 2þ C

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

healthy controls trial; grouped by ages, stratified; questionnaires; peptiduria ASD subgroups
conducted in 2 phases
Mulloy et al, 201161 Systematic review Confirmation of the conclusions 1þ A
reached by Whiteley et al (2010)
Harland 201257 In the recruitment phase, Experimental; double blind; cross Behavioral assessment (activity, Not yet complete _ _
estimated number patients over; RCT; sleep, behaviors related to the
with ASDs 30; 18 wk on GFCF and normal diet, ASD); stool pattern; nutrition
aged between 2 and 17 y alternately
Abbreviations: AA: amino acids; ADHD, attention-deficit/hyperactivity disorder; ASD, autism spectrum disorder; CF, casein free; GF, gluten free; GFCF, gluten-free, casein-free; GR, grade of recommendation; Ig,
immunoglobulin; LE, level of evidence; RCT, randomized controlled trial.
a
The studies classified as 1– or 2– were not used in the recommendation process given the high risk of bias.
 Table 4. Studies Considering the Safety of the Gluten-Free, Casein-Free Diet.

Author N Design Assessment criteria Result LEa GR

Cornish et al, 200211 Cases: 8 children with ASDs, 3-16 y old, GFCF Retrospective, case-control 3-d food diary No differences observed for 2–
diet; controls: 29 children with ASDs, 3-16 y energy and macro- and
old with no dietary restrictions micronutrients
Black et al, 200244 Cases: 50 children who refuse milk (CF), Observational, cross- 4-d food diary; food frequency Lower intake of calcium; stature 2–
controls 200 children who drink milk. Age sectional questionnaire; measurement and bone density associated
range in both groups: 3-10 y of bone density with CF diet

Arnold et al, 200337 Cases: 36 children with ASDs, 26 with normal Observational, cross- Blood analysis Significant deficiency in AA 2þ C
diet, 10 with GFCF diet; controls: 24 sectional (mainly tryptophan) associated
without ASD, both groups under 5 y of age with ASDs compared to
controls, more pronounced
on a GFCF diet
Monti et al, 200745 One 8-y-old child Case report Dairy product elimination diet Lower bone density 3 D
Konstantynowicz et al, Cases: 91 children with fractures; controls: Observational; grouped by 24-hour recall; bone density Lower calcium intake and bone 2þþ B
200746 273 children without fractures; age range: age and sex; randomized; density in children with
2.5-20 y case-control. restriction of dairy products;
weak association between
fractures and milk
consumption
Hediger et al, 200850 75 children with ASDs on a CF diet; no Observational; cross- Cortical bone density Reduced 2þþ B
controls sectional.

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

Johnson et al, 201136 22 children with ASDs aged 3-5 y; cases: 8 Prospective, randomized, 24-h recall No significant differences for 2–
children on a GFCF diet for 3 mo; controls: open-label cohort trial proteins, lipids, carbohydrates,
14 children with no dietary restrictions calcium, and iron
Neumayer et al, 201251 Cases: 18 children with ASD; controls: 19 Observational; case-control Bone mineral density; analysis Lower bone density; lower 2þ C
typically developing children; age range in of blood and saliva; physical physical activity; lower intake
both groups 8-14 y exercise; 3-d food diary of lactose, vitamin D, increase
of salivary cortisol
Abbreviations: AA, amino acids; ASD, autism spectrum disorder; CF, casein-free diet; FFQ: Food frequency questionnaires; GF, gluten-free diet; GFCF, gluten-free, casein-free diet; GR, grade of recommendation; LE, level
of evidence; RCT: randomized clinical trial.
a
The studies classified as 1– and 2– were not used for establishing the recommendations given their high risk of bias.

7
8 Journal of Child Neurology

treatments provided, or the use of 1 or more intervention procedures postintervention comparisons. There was also a risk of bias in
that could affect the assessment criteria selected. data on the behavioral variables attributable to memories of
For evaluating and synthesizing the scientific evidence, we also parents and other caregivers being distorted over time and that
considered the internal validity of the studies, whether there was sta- their perception of changes in the behavior of participants can
tistical significance and the accuracy of the results, as well as their
be subjectively influenced by the fact of being included in non-
clinical relevance. We then characterized the recommendations on the
blinded trials. Similarly, a placebo effect could have had an
basis of the quantity, generality, and clinical relevance of the results as
well as the quality of the scientific evidence. impact on the results. Lastly, alternative explanations were not
always considered, such as the risk of confounding bias, in par-
ticular, it being possible that behavioral improvements were
Results due to ongoing development and behavioral therapy given,
rather than to gluten-free, casein-free diets per se. Finally, it
The studies retrieved were analyzed in terms of the following
should be noted that the literature search cannot have identified
characteristics, as applicable: sample size, study design, assess-
all the relevant publications, and the review itself can be sensi-
ment and intervention criteria, and the results, as well as the
tive to information bias.
level of evidence and the grade of recommendation. Tables 3
and 4 summarize the characteristics of the studies, the results
of which have been referred to above. Recommendations
On the basis of this review, we conclude that the evidence to sup-
Effectiveness port gluten-free, casein-free diets is limited and weak, such diet-
ary restrictions being associated with social rejection,
Scientific literature on this topic is relatively scarce. Among
stigmatization, deficits in socialization and integration, and a
the studies that refer to effectiveness, only four12,34,35,56 can
misuse of resources, as well as potential adverse biomedical
be considered to provide high scientific evidence. The studies
effects. Hence, we advise against resorting to elimination diets
of Harland57 and Hyman58 are not yet completed. Millward59
in an attempt to treat autism spectrum disorders. Specifically,
and Mulloy60,61 present systematic reviews evaluated with the
until there is conclusive evidence of the benefits of gluten-
highest level of evidence and grade of recommendation. Nota-
free, casein-free diets in autism spectrum disorders, they should
bly, in our analysis, the studies that reported positive results
only be introduced after the diagnosis of an intolerance or allergy
were classified with the lowest levels of evidence, whereas the
to allergens in the foods that would be eliminated in such a diet.
rest reported negative results with regards to this type of dietary
Similarly, the results retrieved do not support the opioid theory.
intervention. None of the studies identified provided conclu-
sive evidence because they had poor validity (Table 3).
Implications for the Practice
Safety As a final recommendation, we underline that, when used,
elimination diets must be at least as closely monitored as other
There are similarly few publications addressing the safety of the
types of intervention, to allow doctors, parents, and other care-
gluten-free, casein-free diet. Among those identified, the studies
givers to optimize treatments and hence health outcomes for
of Konstantynowicz et al46 and Hediger et al50 provide the high-
these children. On the other hand, a diet-related specific end
est level of evidence. Nevertheless, in the results found, there
phenotype can be a target for future research and even a marker
was a certain degree of consensus on the risks that could be asso-
for the gluten-free, casein-free dietary intervention.
ciated with following this type of restriction diet (Table 4).
Based on the results of this review, future research should be
focussed on blinded randomized controlled trials, and include
Discussion larger samples sizes.
Data in the literature in this field are very limited both in quan- Author Contributions
tity and quality. To assess the effectiveness and safety of the
All 5 coauthors of this paper have contributed significantly to the
gluten-free, casein-free elimination diet, we considered both
design and implementation of the study, as well as the analysis and
behavioral (verbal and nonverbal communication, stereotypy, interpretation of the results. Further, all have participated in the pre-
and disruptive behavior) and biomedical variables (eg, urinary paration of this manuscript and have approved the final version sub-
peptides, gliadin and endomysial antibodies, as well as other mitted for publication.
laboratory data and nutrient intake). Methodologic limitations
identified were associated with a range of factors: the lack of Declaration of Conflicting Interests
a control group and/or clear definitions of inclusion criteria, The authors declared no potential conflicts of interest with respect to
very small sample sizes, and analysis being based on single the research, authorship, and/or publication of this article.
individuals or anecdotal information, groups being heteroge-
neous in terms of age, failure to control for phenotypic Funding
variability between individuals, interventions being of variable The authors received no financial support for the research, authorship,
duration and generally short, as well as lack of preintervention- and/or publication of this article.

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

Marı´-Bauset et al 9

Ethical Approval tandem mass spectrometry method to test the opioid excess the-
The study protocol was approved by the Ethics Committee of the ory. Anal Bioanal Chem. 2007;388:1643-1651.
University Hospital ‘‘Dr. Peset’’ (Valencia, Spain) (code 46/10). The 20. Cass H, Gringras P, March J, et al. Absence of urinary opioid pep-
school accepted the study, and parents of the children participating in tides in children with autism. Arch Dis Child. 2008;93:745-750.
the study gave written informed consent. 21. Wong HH, Smith RG. Patterns of complementary and alternative
medical therapy use in children diagnosed with autism spectrum
References disorders. J Autism Dev Disord. 2006;36:901-910.
1. Filipek PA, Accardo PJ, Baranek GT, et al. The screening and diag- 22. Herndon AC, DiGuiseppi C, Johnson SL, et al. Does nutritional
nosis of autism spectrum disorders. J Autism Dev Disord. 1999; intake differ between children with autism spectrum disorders and
29(6): 439-484. Review. Erratum in: J Autism Dev Disord. 2000; children with typical development? J Autism Dev Disord. 2009;
30: 81. 39:212-222.
2. CDC. Centers for Disease Control and Prevention. Prevalence of 23. Bandini LG, Anderson SE, Curtin C, et al. Food selectivity in
autism spectrum disorders. MMWR Surveill Summ. 2012;61:1-18. children with autism spectrum disorders and typically developing
Available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/ children. J Pediatr. 2010;157:259-264.
ss6103a1.htm. Accessed September 11, 2013. 24. Hall SE, Riccio CA. Complementary and alternative treatment
3. Baghdadli A. Recommandations pour la pratique professionnelle use for autism spectrum disorders. Complementar Ther Clin
du diagnostic de l’autisme. Fédération française de psychiatrie et Pract. 2012;18:159-163.
haute autorité de santé. 2005. Available at: http://has-sante.fr/por- 25. Sharp WG, Jaquess DL, Lukens CT. Multi-method assessment of
tail/display.jsp¼c468812. Accessed September 30, 2013. feeding problems among children with autism spectrum disorders.
4. Francis K. Autism interventions: a critical update. Dev Med Child Res Autism Spectr Disord. 2013;7:56-65.
Neurol. 2005;47:493-499. 26. Dohan FC.Cereals and schizophrenia. Data and hypothesis. Acta
5. Levy SE, Hyman SL. Novel treatments for autistic spectrum dis- Psychiatr Scand. 1966;42:125-152.
orders. Ment Retard Dev Disabil Res Rev. 2005;11:131-142. 27. Dohan FC, Graberger JC. Relapsed schizophrenics: earlier dis-
6. Harrington JW, Rosen L, Garnecho A, Patrick PA. Parental per- charge from the hospital after cereal-free, milk-free diet. Am J
ceptions and use of complementary and alternative medicine Psychiatry. 1973;130:685-686.
practices for children with autistic spectrum disorders in private 28. Panksepp JA. A neurochemical theory of autism. Trends Neu-
practice. J Dev Behav Pediatr. 2006;27:156-161. rosci. 1979;2:174-177.
7. Hanson E, Kalish LA, Bunce E, et al. Use of complementary and 29. Knivsberg AM, Wiig K, Lind G, et al. Dietary intervention in
alternative medicine among children with autism spectrum disor- autistic syndromes. Brain Dysfunction. 1990;3:315-327.
der. J Autism Dev Disord. 2007;37:628-636. 30. Reichelt KL, Knivsberg AM, Lind G, Nødland M. Probable etiol-
8. Alpert M. The autism diet. Sci Am. 2007;296:19-20. ogy and possible treatment of childhood autism. Brain Dysfunc-
9. Whiteley P, Rogers J, Savery D, Shattock P. A gluten-free diet as tion. 1991;4:308-319.
an intervention for autism and associated disorders: preliminary 31. Knivsberg AM, Reichelt KL, Høien T, Nødland M. Effect of diet-
findings. Autism. 1999;3:45-65. ary intervention on autistic behavior. Focus Autism Other Dev
10. Cade R, Privette M, Fregly M, et al. Autism and schizophrenia: Disabl. 2003;18:247-256.
intestinal disorders. Nutr Neurosci. 2000;3:57-72. 32. Reichelt KL, Knivsberg AM. Can the physiopathology of autism
11. Cornish E. Gluten and casein free diets in autism: a study of the be explained by the nature of the discovered urine peptides? Nutr
effects on food choice and nutrition. J Hum Nutr Diet. 2002;15: Neurosci. 2003; 6: 19-28.
261-269. 33. Whiteley P, Shattock P, Knivsberg AM, et al. Gluten and casein-
12. Sponheim E.Gluten-free diet in infantile autism. A therapeutic free dietary intervention for autism spectrum conditions. Front
trial. Tidsskr Nor Laegeforen. 1991;111:704-707. Hum Neurosci. 2013;6:344.
13. Amidon GL, Lee HJ. Absorption of peptide and peptidomimetic 34. Elder JH, Shankar M, Shuster J, et al. The gluten-free, casein-free
drugs. Annu Rev Pharmacol Toxicol. 1994;34:321-341. diet in autism: results of a preliminary double blind clinical trial.
14. Daniel H. Molecular and integrative physiology of intestinal pep- J Autism Dev Disord. 2006;36:413-420.
tide transport. Annu Rev Physiol. 2004;66:361-384. 35. Seung HK, Rogalski Y, Shankar M, Elder J. The gluten-and
15. Janecka A, Staniszewska R, Gach K, Fichna J. Enzymatic degra- casein-free diet and autism: communication outcomes from a pre-
dation of endomorphins. Peptides. 2008;29:2066-2073. liminary double-blind clinical trial. J Med Speech Lang Pathol.
16. Zioudrou C, Streaty RA, Klee WA. Opioid peptides derived from 2007;15:337- 345.
food proteins. The exorphins. J Biol Chem. 1979;254:2446-2449. 36. Johnson CR, Handen BL, Zimmer M, et al. Effects of gluten free-
17. Lennernäs H. Intestinal permeability and its relevance for absorp- casein free diet in young children with autism: a pilot study. J Dev
tion and elimination. Xenobiotica. 2007;37:1015-1051. Phys Disabil. 2011;23:213-325.
18. Sponheim E, Myrhe AM, Reichelt KL, Aalen OO. Urine peptide 37. Arnold GL, Hyman SL, Mooney RA, Kirby RS. Plasma amino
patterns in children with milder types of autism. Tidsskr Nor Lae- acids profiles in children with autism: potential risk of nutritional
geforen. 2006;25:1475-1477. deficiencies. J Autism Dev Disord. 2003;33:449-454.
19. Dettmer K, Hanna D, Whestone P, et al. Autism and urinary exo- 38. Thompson T. Folate, iron, and dietary fiber contents of the gluten-
genous neuropeptides: development of an on-line SPE-HPLC- free diet. J Am Diet Assoc. 2000;100:1389-1396.

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

10 Journal of Child Neurology

39. Hallert C, Grant C, Grehn S, et al. Evidence of poor vitamin status 57. Harland SW. A study to assess the role of a gluten free-dairy free
in celiac patients on a gluten-free diet for 10 years. Aliment (GFCG) diet in the dietary management of autism associated gas-
Pharmacol Ther. 2002;16:1333-1339. trointestinal disorders. Clinical Trials.gov. Available at: http://
40. Ali A, Waly MI, Al-Farsi YM, et al. Hyperhomocysteinemia clinicaltrials.gov/ct2/show/NCT01116388?term¼autismþdiet&
among Omani autistic children: a case-control study. Acta rank¼2. (NCT01116388). 2010. Accessed September 11, 2013.
Biochim Pol. 2011;58:547-551. 58. Hyman SL.Diet and behavior in young children with Autism.
41. Mariani P, Viti MG, Montuori M, et al. The gluten-free diet: a Clinical Trials.gov. Available at: http://clinicaltrials.gov/ct2/
nutritional risk factor for adolescents with celiac disease? show/NCT00090428?term¼autismþdiet&rank¼2. (NCT000904
J Pediatr Gastroenterol Nutr. 1998;27:519-523. 28). Accessed September 05, 2013.
42. Marcason W. What is the current status of research concerning 59. Millward C, Ferriter M, Calver S, Connell-Jones G. Gluten- and
use of a gluten-free, casein-free diet for children diagnosed with casein-free diet for autistic spectrum disorder. Cochrane Data-
autism? J Am Diet Assoc. 2009;109:572. base Syst Rev. 2008; 16 (2): CD003498. Available at: http://
43. Marcason W. Is there evidence to support the claim that a gluten-free www.update-software.com. Accessed September 16, 2013.
diet should be used for weight loss? J Am Diet Assoc. 2011;111: 60. Mulloy A, Lang R, O’Reilly M, et al. Gluten-free and casein-free
1786. diets in the treatment of autism spectrum disorders: a systematic
44. Black C, Kaye JA, Jick H. Relation of childhood gastrointestinal dis- review. Res Autism Spectr Disord. 2010;4:328-339.
orders to autism: nested case-control study using data from the UK 61. Mulloy A, Lang R, O’Reilly M, et al. Addendum to ‘‘gluten-free
General Practice Research Database. BMJ. 2002;325:419-421. and casein-free diets in treatment of autism spectrum disorders: a
45. Monti G, Libanore V, Marinaro L, Lala R, Miniero R, Savino F. systematic review.’’ Res Autism Spectr Disord. 2011;5:86-88.
Multiple bone fracture in an 8 years old child with cow’s milk 62. Goodwin MS, Cowen MA, Goodwin TC. Malabsorption and cer-
allergy and inappropriate calcium supplementation. Ann Nutr ebral dysfunction: a multivariate and comparative study of autis-
Metab. 2007;51:228-231. tic children. J Autism Child Schizophr. 1971;1:48-62.
46. Konstantynowicz J, Nguyen TV, Kaczmarski M, et al. Fractures 63. Bird BL, Russo DC, Cataldo MF. Considerations in the analysis
during growth: potential role of a milk-free diet. Osteoporos Int. and treatment of dietary effects on behaviour: a case study. J
2007;18:1601-1607. Autism Child Schizophr. 1977;7:373-381.
47. Agostoni C, Turck D. Is cow’s milk harmful to a child’s health? J 64. O’Bannion D, Armstrong B, Cummings RA, Stange J. Disruptive
Pediatr Gastroenterol Nutr. 2011;53:594-600. behaviour: a dietary approach. J Autism Child Schizophr. 1978;8:
48. Aldámiz-Echevarria L, Bilbao A, Andrade F, et al. Fatty acid 325-337.
deficiency profile in children with food allergy managed by elim- 65. McCarthy DM, Coleman M. Response of intestinal mucosa to
ination diet. Acta Paediatr. 2008;97:1572-1576. gluten challenge in autistic subjects. Lancet. 1979;2:877-878.
49. Schuchardt JP, Huss M, Stauss-Grabo M, Hahn A. Significance of 66. Reichelt KL, Ekrem J, Scott H. Gluten, milk proteins and autism:
long-chain polyunsaturated fatty acids (PUFAs) for the develop- dietary intervention effects on behaviour and peptide secretion. J
ment and behaviour of children. Eur J Pediatr. 2010;169:149-164. Appl Nutr. 1990;42:1-11.
50. Hediger ML, England LJ, Molloy CA, et al. Reduced bone corti- 67. Lucarelli S, Frediani T, Zingoni AM, et al. Food allergy and
cal thickness in boys with autism or autism spectrum disorder. J infantile autism. Panminerva Med. 1995;37:137-141.
Autism Dev Disord. 2008;38:848-856. 68. Knivsberg AM, Reichelt KL, Nødland M, Høien T. Autistic syn-
51. Neumeyer AM, Gates A, Ferrone C, et al. Bone density in peripu- dromes and diet: a follow-up study. Escanear Educ Res. 1995;
bertal boys with autism spectrum disorders. J Autism Dev Disord. 39:223-236.
2013;43:1623-1629. 69. Adams L, Conn S. Nutrition and its relationship to autism. Focus
52. Friedenreich CM. Methods for pooled analyses of epidemiologic Autism Other Dev Disabl. 1997;12:53-58.
studies. Epidemiology. 1993;4:295-302. 70. Knivsberg AM, Reichelt KL, Nødland M. Dietary intervention for a
53. Froom P, Froom J. Deficiencies in structured medical abstracts. J seven year old girl with autistic behaviour. Nutr Neurosci. 1999;2:
Clin Epidemiol. 1993;46:591-594. 435-439.
54. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement 71. Knivsberg AM, Reichelt KL, Høien T, Nødland M. A rando-
for reporting systematic reviews and meta-analyses of studies that mized, controlled study of dietary interventions in autistic syn-
evaluate healthcare interventions: explanation and elaboration. dromes. Nutr Neurosci. 2002;5:251-261.
BMJ. 2009;339:b2700. 72. Irvin D. Using analog assessment procedures for determining the
55. Scottish Intercollegiate Guidelines Network. A guideline effects of a gluten-free and casein-free diet on rate of problem beha-
developers’ handbook (Publication no. 50). Edinburgh: SIGN: viors for an adolescent with autism. Behav Interv. 2006;21:281-286.
2001 (updated in January 2008). Available at: http://www.sign. 73. Patel K, Curtis LT. A comprehensive approach to treating autism
ac.uk/guidelines/fulltext/50/index.html. Accessed September and attention-deficit hyperactivity disorder: a prepilot study. J
30, 2013. Altern Complement Med. 2007;13:1091-1097.
56. Hyman S, Stewart P, Smith T, et al. The gluten free and casein free 74. Whiteley P, Haracopos D, Knivsberg AM, et al. The ScanBrit
(GFCF) diet: A double blind, placebo controlled challenge study. randomised, controlled, single-blind study of a gluten- and
Paper presented at the International Meeting for Autism Research, casein-free dietary intervention for children with autism spectrum
Philadelphia, PA 2010. disorders. Nutr Neurosci. 2010;13:87-100.

Downloaded from jcn.sagepub.com at Universidad de Valencia on May 5, 2014

View publication stats