Journal of Autism Developmental Disorders, Research Paper Example

The Gluten Free-Casein Free Diet in Autism: Results of a Preliminary Double Blind Clinical Trial

Authors: Elder, H.E., Shankar, M., Shuster, J., Theriaque, D., Burns, S.& Lindsay, S. (2006).Source: Journal of Autism Developmental Disorders Vol. 36 No.3 pp.413-420

Analysis of Literature Review

Is there adequate rationale

The paper was about a preliminary study on the efficacy of a Gluten Free-Casein Free (GFCF) diet in children diagnosed with autism over a twelve week period, using the double blind repeated measure crossover design, and was conducted by Elder, Shankar, Schuster, Burns &Lindsay (2006).

However, the rationale for dietary treatment of child behavior disorders were not sufficiently justified by Elder et al. (2006), with respect to the data provided, as they only used the work Feingold (1975) which was performed in a cultural environment very different from they operated in, and where all experimental controls and limitations were not supplied for proper comparisons to be made.

No explanations as to why only one twelve week period of testing was conducted, rather than at least three  which would enable the growth of their learning curves regarding conducting clinical test on GFCF diets to autism diagnosed children, as well as to effect timely corrective measures.

These shortcomings therefore reflected inadequacies in the rationale regarding the Literature presented as the basis for conducting the study.

Are all important concepts defined?

The authors failed to properly define Autism at the crucial introductory stage of the study, as well as to justify the reasons why they claim the disease was increasingly prevalent and devastating to families.

Information regarding the use of the double blind test theory, using Hover and Millch (2006) work, was a commendable strategy by Elder et al. (2006), in that it provided evidence of the effectiveness and reliability of the process.

However, there was no proper definition and communication of this very important concept to the audience including parents regarding well as how it came to be the preferred approach in scientific studies of this nature, bearing in mind that the lack of knowledge of the technique on the part of the parents allowed errors to be introduced into the study.

Behavioral patterns associated with the disorder, inclusive of limited reciprocal interactions, marked impairments and functional communication abilities, however, were successfully communicated, as well as the  mention of numerous theoretical models in the scientific community that were providing biological explanations for the causes or existence of autism.

There was a significant time gap between 1920 and 1970, where Elder et al. (2006) failed to mention if any studies were conducted that might influence their  conceptual development and application as well as methods of approach to the study being conducted.

Credit however, should be given to the researchers for the detailed and precise explanations given concerning their statistical tools and measuring instruments used in the study.

Is there potential for difference?

In terms of the sources to justify the choice of a Gluten Free-Casein Free diet, Elder et al. (2006), provided only one, which was inadequate in terms of objectivity and authenticity which are vital to gaining approval in the scientific community. Additionally, the argument by Elder et al. (2006), that there inadequate empirical evidence supporting the use of GFCF diets to cure autism, points to the possibility of differences in expected outcomes.

Are there biases evident?

The authors exemplified the narrowness of their consultancy efforts, by mentioning that that they only worked collaboratively with the University Of Florida General Clinic Research Center, to overcome a number of methodical issues and had also incorporated findings from their prior NINR.NH funded studies concerning GFCF ASD related studies, without specifying what findings were extracted. It could be argued that Elder et al. (2006) were bias in their selection by this approach, and this could allow defects in past studies being replicated in the study and potentially reduce the accuracy on data compiled and analyses conducted.

Is it thorough and comprehensive?

The authors should be commended for including the suggestion of Wolraich (1996) in the literature. Wolraich (1996), warned researchers who were engaged in dietary implementations, to exercise caution and observe the restrictions that accompany the diets that were being promoted. The move could be regarded as an attempt to introduce balance, control and protection for the children who would be taking part in future preliminary studies.

However, the researchers also failed to adequately quantify the number of early research works that had recorded the adverse effects of sugar on the hyperactive and aggressive behaviors manifesting among children with learning disability, when they only mentioned Prinz, Roberts & Hantman (1980) and Wolraich, Stumbo & Millch (1986).

Are all relevant paradigms discussed?

The literature also had no report on scientific work done with parents playing any role similar to the role they played in the study conducted by Elder et al. (2006), and hence there was no control or methodology of approach as to how they should operate during the research.  It was not clear why the researchers used this approach and they fail explained this very important paradigm and its consequences on the study.

Additionally, Elder et al. (2006) use of Feingold (1975) work without elaborating the reasons for the researcher achieving only a 50% success rate among hyperactive and disabled children, who were placed on diets free of salicylate and additives, showed a lack of comprehensiveness in the literature.

The criticism is justified,in that Elder et al (2006) reported that other researchers had used Feingold (1976) work among others, as base to promote dietary treatments, by restricting additives, sugars and preservatives. Had Elder et al. (2006), taken this approach on page 414 of the Literature Review, the study would  no doubt be deemed more balance and objective in the scientific community, in addition to the opportunity they would have in making timely adjustments to their procedural approach.

Is there clarity?

The literature lacked clarity from the very beginning, when the disease was not properly defined, followed by the gap of nearly fifty years of no reported work on the disease, the lack of reason for using Feingold (2006) 50% success rate work, the choice of such small, age disparate, gender imbalanced and practically homogenous sample size, the reason for the use of scientifically unsupervised parents, as well as the ambiguity in the purpose of the study..

Finally it was not clear why Elder et al. (2006)   mentioned in the literature that some parents believe that dietary restrictions greatly affected their children behavior, although there were no scientific evidence to support such a statement was improper, and a literature review should not be a forum to include laymen arguments or opinions, but for the presentation of scientifically verifiable information that emphatically support studies being done.

Analysis of Purpose Statement

Is one clearly evident

The real purpose statement of the study was not evident at the outset of the paper, but was inferred that it was to demonstrate how to conduct control clinical trials of GFCF diets, as well as to provide preliminary work suggestions and recommendations for directing future research.

Elder et al (2006)mentioned later in the study that the purpose was to provide critical feedback loop for families of children with ASD, including the most current scientifically sound information, and incorporating family observations into treatment and planning development. However, at the end the study there was nothing of significance to communicate to families to solve their autistic problems that was evident among their children. This purpose was not achieved.

Is there a clear connection to the literature review?

However, evidence of the connection of the unstated real purpose of the study being connected to the literature, was observed in the first paragraph on page 414 of the paper, and it stated that the study not only yielded preliminary data regarding diet efficacies, but also provided important information on developing and implementing GFCF diets as well as clarifying a number of feasibility issues.

Are variables clearly indentified?

Uses made of the Likert Scale to measure and appropriate responses, the double blind concept, and control and sampling techniques, were variables that served to connect the purpose statement of the preliminary study with the contents of the literature.

Is the target population clearly identified?

Elder et al. (2006) had also clearly established in the literature that the 15 children who were diagnosed with suffering from autism, as the target population for the study in which the efficacy of the GFCF diet would be applied over the 12 weeks.

Is the purpose clearly researchable?

The purpose stated by Elder et al. (2006),was not clearly researchable, as the age difference within the targeted populationwas too wide; with the youngest being 2 years and the oldest at 16 years. This gap may have significantly affected the consistency and measurement of the behavioral responses within the study, in that the level of food intake and response times needed may vary considerable, and there might be no parameters to handle and compensate for these variations.

Additionally, the sample size of 15 which consist of 12 males and 3 females as well as the 96 % Caucasian composition, robs the study of the opportunity generate scientific data that could be classified as balanced in gender and diversity.

Analysis of the Hypothesis

What is the type of hypothesis?

Elder et al. (2006), study was a preliminary research type hypothesis that was directional and testable. The research team stated emphatically at the outset, that the study was a pilot project that was designed to test the efficacy of a Gluten Free and Casein Free (GFCF) diet among autism diagnosed children, using a double blind repeated measure cross over approach for a period of twelve weeks.

Does the literature review justify directionality?

The hypothesis grew out of the work of Cade et al. (2000), who conducted a series of tests based on the hypothesis that schizophrenia was in some way associated with the absorption of exorphins contained in gluten and casein.

Elder et al. (2006), by using Cade et al (2000) work as the basis for their hypothesis, together with the reported low levels of schizophrenia from diets that excluded wheat, rye, barley and oats in the New Guinea and South Pacific population, and the perceived threat from the same food causing autism in children today, were ensuring that the hypothesis for their study had directionality and could achieve similar success.

Is the hypothesis testable and falsifiable?

Their hypothesis was testable due to the fact that the researchers had identified a relevant target population using a scientific sampling process and were incorporating a double blind approach which was confirmed as ideal for the study during the clinical trials.

Additionally, Elder et al. (2006) were also applying the reputable Childhood Autism Rating Scale (CARS), and the Autism Diagnostic –Revised (ADI-R) instruments to conduct their measurements, along with Urinary Peptide Levels (UPL) tests and Ecological Communication (ECO) Language Sample Summary to record child behaviors and collect interactive samples.

On the other hand, the hypothesis was falsifiable for several reasons, and they include the small sample size, the structural composition of the sample in terms of age distribution, gender and low diversity ratio, the inclusion of parents to play a vital role in a scientific experiment that they were not properly trained and constantly supervised.

Furthermore, apart from the University of Florida GCRC Bio-nutrition staff providing the meals requirements, the target population could have been supplied with supplemental foods from a preferential list without the presence of any coordinators for the study. This could haveinduced errors into the readings obtained.

The timeline for the study was also a limiting factor on the accuracy and reliability of the results obtained, in that the huge gap in the age composition of the group and the possible different consumption and reaction rates, could make the differences between readings insignificant and difficult to rationalize.

Had Elder et al. (2006) employed a larger population, with members with more evenly distributed subgroup in terms of age, gender and race, and instituted longer periods for observation, significant results could have been achieved, as there would be more consistency in the data generated as well as effective comparisons in the results.

In paragraph 3 on page 419 of the article, according to Elder (2006), there were some indications that the labels were not read carefully during the clinical trials. Soy products according to Elder et al (2006), may have been mistaken for urinary peptides and allowed the introduction of confounds in to the digestive systems of the children. This also would negatively impact the outcome of the trials.

There was also further evidence that the hypothesis wouldnot be properly tested, when Elder et al. (2006) chose a small sample size yet failed to manage the application process, to ensure the hypothesis could be tested under properly established conditions. They alsoassume that the parents’ literacy levels were at the required standard to participate in a clinical trial of this nature. These weaknesses imposed severe limitations on the accuracy of the results that would be generated.

Reasons for falsifications were also evident from the fact that Elder et al., (2006) used Cade et al (2000) work as the basis for their hypothesis, but failed to provide more details on the study conducted in New Guinea and the Pacific Islands, especially regarding what part of that study was replicated in theirs and what were the limitations during Cade et al. (2000) trials.

Readers would have wanted to know if the members of the two islands were subjected to the same operating conditions like the children with autism, in terms of the food supplied, the time framesand whetherthe populations in New Guinea and the Pacific Islands may have ingested supplemental foods similar to or unlike those of the children diagnosed with autism.Elder et al. (2006) in this regard provided too much sources of possible errors in the study and this should surely lead to falsifications in the hypothesis, as well as questions regarding the integrity of the program.

Is the hypothesis reasonable?

The researchers, for reasons best known to them, conducted only one twelve week period of study to test the hypothesis, and such provided no other opportunities to make proper scientific comparisons and evaluations of data generated from similar operating procedures, so that medications and adjustments could be made to furthers fine tune the clinical trials, and ensure that results obtained would have a greater probability to be considered significant.

Elder et al. (2006) under close examinationshad a number of shortfalls and inadequacies regarding the literature review, which was presented and as such there are valid questions regarding whether the study and the hypothesis should be considered reasonable.

The team of scientists failed to (a) properly define the disease because they may have assumed the readers already knew, and (b) to present clear pictures of what the purpose of the trials were about, when they mentioned two different purposes at beginning of the study and on page 415.

They also lacked proper connections between their stated purpose statement and the literature presented, and may have unknowingly or knowingly misled their reading audience as to the real purpose of the preliminary study.

It could be argued within reason, which the purpose of the study was to provide a demonstration of how to conduct clinical trials using of GFCF diet using the double blind procedure, as well as to provide guidelines and directions for future research, because there was nothing substantive to report to the parents at the end of the study.

Serious questions must also be asked about the purposeful sampling process conducted by the Center for Autism and Related Disabilities or Child Psychiatry Services at the University of Florida Department of Psychiatry and Brain Institute. How could a reputable institution of this nature recommend such a small sample size to be used in a critical experiment of this nature? Why were there not other samples developed and used simultaneously?

The reason for Elder et al. (2006) to only consult this institution for sampling and not contacting others is also questionable, as other states and institutions could have been able to provide other target populations that would be appropriate for this study and facilitate synchronicity in the study.

It seems that they researcher who were all based at Gainesville in the University of Florida may have had had logistic problems or were limited in their visions to accept this severely deficient sampling to conduct trials, from which results were expected to be used to advise families with autism related children with significantly different characteristics than what prevailed in the study. This approach should be considered narrow, subjective and somewhat unprofessional for an experiment of such national importance.

It could also be deemed unreasonable for Elder et al. (2006) to employ the services of the parents in the trials without any precedent being established in the Literature Review by other notable researchers like Cade et al. (2000), Hastings &Brown (2002), and Dohan (1966). Elder et al (2006) had nothing to base this aspect of the operation and made no mention of it even in their reporting on the limitations of the study.

There were also no reports prior to the study of any kinds of evaluations done to these parents, especially psychologically, physically and intellectually, before they were asked to play such vital role in a trial that require zero bias and complete adherence to the stated procedure.

Is the hypothesis parsimonious?

The hypothesis presented by Elder et al. (2006) could be described as parsimonious, in that the information presented in the literature was inadequate for the purpose given for the study, despite the fact that their overall communications were quite clear. It could be deemed intuitively that the research team itself lacked proper intellectual guidance and supervision, even after having completed two other autism related trials. They also were not giving adequate considerations to the type of audience they would be communicating with, and what were the expectations, based on their failure to provide more details and proper definitions on the past autism research that were done

Rather than seeking fresh ideas regarding the GFCF diet to supplement the knowledge they gained from their three previous NINR/NIH funded related studies on ASD children, Elder et al. (2006) have deprived the study of much needed unbiased and competent external sources, which would have enabled the hypothesis to be properly tested and not be considered parsimonious in quality at the end.

Is the hypothesis clear and succinct?

The goals that Elder et al (2006), wanted to achieve were clear and succinct, in that they wanted to conduct a clinical trial using GFCF diet among autism diagnosed children under control double blind conditions, and then report their findings to parents and the wider academia, but they went about it without fully equipping themselves with the knowledge and levels of expertise required, as well with a hypothesis was connected to the literature and the procedure applications.

Reference

Elder, H.E., Shankar, M., Shuster, J., Theriaque, D., Burns, S. & Lindsay, S. (2006).The Gluten Free-Casein Free Diet in Autism: Results of a Preliminary Double Blind Clinical Trial

Journal of Autism Developmental Disorders Vol. 36 No.3 pp.413-420