Ratings of the Effectiveness of Nutraceuticals for Autism Spectrum Disorders: Results of a National Survey

Autism spectrum disorder (ASD) often involves a wide range of co-occurring medical conditions (“comorbidities”) and biochemical abnormalities such as oxidative stress and mitochondrial dysfunction. Nutritional supplements (“Nutraceuticals”) are often used to treat both core ASD symptoms and comorbidities, but some have not yet been formally evaluated in ASD. The potential biological mechanisms of nutraceuticals include correction of micronutrient deficiencies due to a poor diet and support for metabolic processes such as redox regulation, mitochondrial dysfunction and melatonin production. This paper reports on the results of the National Survey on Treatment Effectiveness for Autism, focusing on nutraceuticals. The Survey involved 1286 participants from across the United States. Participants rated the overall perceived benefits and adverse effects of each nutraceutical, and also indicated the specific symptoms changed and adverse effects. From these ratings the top-rated nutraceuticals for each of 24 symptoms are listed. Compared to psychiatric and seizure medications rated through the same Survey, on average nutraceuticals had significantly higher ratings of Overall Benefit (1.59 vs. 1.39, p = 0.01) and significantly lower ratings of Overall Adverse Effects (0.1 vs. 0.9, p < 0.001). Folinic acid and vitamin B12 were two of the top-rated treatments. This study suggests that nutraceuticals may have clinical benefits and favorable adverse effect profiles.


Introduction
Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder involving core problems in social communication and repetitive behaviors and affects about 2% of children in the United States [1]. A number of medical conditions cooccur with ASD (termed "comorbidities"), including intellectual disability [2], epilepsy [3], gastrointestinal disorders (such as constipation and diarrhea) [4], sleep disorders [5], attention deficit disorder [5], anxiety [5], and irritability, self-injurious behavior, and depression [6]. Other studies have reported biochemical abnormalities, including problems with methylation pathway insufficiency [7,8], insufficient production of melatonin for sleep [9], mitochondrial dysfunction [10,11] and oxidative stress [12,13]. Currently, there are no FDA-approved medications for treating the core symptoms of ASD (social communication and restricted/repetitive behaviors), although there are two FDA-approved medications for treating the associated symptom of irritability [14].
Compared to typically developing (TD) children, feeding difficulties are common in children with ASD and include food refusal, eating a limited variety of foods and having more problems with mealtime behavior [15], and they may have nutrient-poor diets [16]. One meta-analysis of 17 prospectively controlled studies reported significantly more feeding problems in children with ASD compared to controls (odd ratios 5.11, 95% CI 3.74-6.97) and significantly lower intake of calcium and protein in the ASD group [17]. A systematic review of 29 studies reported that feeding problems were associated with impaired sensory processing, perception, more rigidity and challenging behaviors [18]. A prospective, randomized controlled trial of a comprehensive dietary and nutritional intervention found that a combination of six treatments (vitamins/minerals, essential fatty acids, Epsom salts, carnitine, digestive enzymes, and a healthy low-allergen diet) led to significant improvements over one year in autism symptoms, developmental age, and non-verbal IQ compared to controls [19].
Because of potential deficiencies of nutrients (often related to feeding problems) and biochemical abnormalities (e.g., oxidative stress, mitochondrial dysfunction, methylation problems, among others) reported in individuals with ASD, a number of studies have investigated the use of vitamins, minerals, and other nutritional supplements (hereafter termed "Nutraceuticals"). The use of nutraceuticals is typically considered a form of complementary and alternative medicine (CAM), although many nutraceuticals are based on the science of nutritional biochemistry and target deficiencies and biochemical problems. Owen-Smith et al. (2015) conducted a survey of 42 CAM treatments used in ASD and reported 88% of participants had been treated with at least one CAM treatment [20]. Frye et al. (2011) surveyed the effectiveness of seizure treatments (including nutraceuticals) in 733 children with ASD and seizures compared to 290 controls and reported some CAM treatments (such as vitamin B6, magnesium, taurine, and vitamin B12) were rated as helpful for treating seizures [21]. In addition, a large online survey was conducted by the Autism Research Institute (the "Parent Ratings of Behavioral Effects of Biomedical Interventions Survey") [22]. This survey of 27,000 parents of individuals with ASD rated the effectiveness of 84 various medications, supplements, and diets, using a six-point scale from "made worse" to "made better"; a number of treatments were reported as beneficial, including methylcobalamin (MB12), melatonin, digestive enzymes, fatty acids, cod liver oil, vitamin B6, zinc, magnesium, folic acid, vitamin C, and vitamin A [23]. Although these surveys focused on the overall effectiveness ratings for medications and nutraceuticals used in ASD individuals, most of these studies did not utilize a separate rating scale for the benefits and adverse effects (AEs) and did not obtain information on the effects of these treatments on specific symptoms of ASD.
Some of the medical comorbidities and biochemical abnormalities reported in individuals with ASD might improve with nutraceuticals. For example, randomized clinical trials for ASD have demonstrated the efficacy of melatonin supplementation [9], folinic acid [24][25][26], vitamin/mineral supplements [27,28], comprehensive nutritional interventions [19], N-acetyl cysteine (NAC) [29], and sulforaphane [30]. This paper presents the results of a national survey (the "National Survey on Treatment Effectiveness for Autism") in individuals with ASD and contains more extensive assessments of the treatment effects on specific behaviors and AEs of nutraceuticals in ASD. A previous paper from this Survey reported on the results for psychiatric and seizure medications [31].

Materials and Methods
The research team created the "National Survey on Treatment Effectiveness for Autism" (from now on referred to as "the Survey") and obtained reviews by families of children/adults with ASD and experts in a variety of fields who treat individuals with ASD. This study was approved by the Institutional Review Board of Arizona State University (STUDY00003766). The Survey was advertised to families of individuals with ASD across the country with the assistance of over 50 autism organizations (see Acknowledgements). A full explanation of the Survey creation and distribution can be found in the previous paper [31]. The Survey obtained general medical history and the use of psychiatric and seizure medications, general medications, nutraceuticals, diets, therapies, and information on Kindergarten through grade 12 education. This paper reports data only on the nutraceutical section from Survey responses from 1710 people (of which 1286 (75.2%) rated the effects of nutraceuticals); additional responses were collected since the analysis reported in the previous paper. The exact diagnosis of the individual with ASD was queried using the following categories: autism, Asperger's syndrome, autism spectrum disorder, high-functioning autism, pervasive developmental disorder not otherwise specified (PDD-NOS), no current diagnosis but was previously on the autism spectrum, and "other" in order to capture both DSM-IV and DSM-5 diagnostic categories. These diagnoses were reported by the participant, but not verified in this study since it was an anonymous survey.
The Survey was divided into sub-sections for various types of nutraceuticals (amino acids, vitamins, etc.). At the beginning of each sub-section, the Survey asked what nutraceuticals the participant had taken (from a list of 123 nutraceuticals found in Table S1). For each nutraceutical taken, the Survey asked the participant to rate the overall perceived benefit of the nutraceutical (no benefit = 0, slight benefit = 1, moderate benefit = 2, good benefit = 3, great benefit = 4), the primary symptoms benefited (if any), the overall AE of the nutraceutical (no adverse effect = 0, mild adverse effect = 1, moderate adverse effect = 2, severe adverse effect = 3), and the specific symptoms that were adversely affected (if any). Table 1 shows the symptom list from which participants could select (they could select one or more for each treatment). Finally, the Survey asked for the overall average effect of all nutraceuticals (on a 7-point scale ranging from "much better" to "much worse"). Only treatments with 20 or more responses were included in this analysis. It should be noted that the ratings are the perceived benefit of the evaluator (primarily a caregiver or sometimes the person with ASD), and not ratings by a clinician or physician, which is a limitation of the study.
For each treatment, the top 3 benefits were reported as well as any other benefits with over 20% of participants reporting a benefit. For AEs, the top 3 AEs were reported and any other AEs which were reported by 15% or more of participants. These were arbitrary cut-offs to limit table entries to the most relevant symptoms; a slightly lower cut-off for AEs was chosen since they were so rare.
The top-rated treatments for each symptom were calculated by multiplying the overall net benefit by the percentage of participants who had improvements in that symptom. For each symptom, the three top-rated treatments are reported, as well as any other treatments with a score of 0.2 or higher (equivalent to 10% of participants reporting a moderate benefit).
In order to determine if any of the nutraceuticals were related to changes in ASD severity, two questions were asked on the Survey. Specifically, the ASD severity rated at 3 years of age (which would be close to most patient's diagnosis) was compared to the currently rated ASD severity. The categories of severity were coded on a five-point scale with increasing numeric values corresponding to increasing severity. Specifically, no symptoms (0), very mild symptoms (1), mild symptoms (2), moderate symptoms (3), and severe symptoms (4). The current ASD severity was subtracted from the severity at baseline (3 years of age) such that a decrease in severity would indicate an improvement. The generalized linear model performed in IBM SPSS PASW Release 18.0.0 (Armonk, New York) was used to analyze change in severity. The model included gender (male, female), developmental profile, number of antibiotic treatments in the first 3 years of life (since that has been reported higher in ASD), and baseline ASD severity. In general, only treatments that were used by 100 respondents or more were analyzed to ensure generalizability and a wide range of ASD severity changes. Two different approaches were used. First, it was determined whether use of the nutraceutical was associated with improvements in ASD symptoms by comparing those who used the nutraceutical to those who did not. Second, the association between the perceived benefits of the nutraceutical with the change in ASD symptoms was examined by comparing current ASD severity versus severity at 3 years old. This later analysis including an interaction between treatment and severity at 3 years of age in order to determine whether the change in severity associated with the treatment was affected by the severity of ASD at age 3 years of age. A one-way analysis of variance was also used to determine whether severity at 3 years of age was related to the use of any treatment studied.

Demographics and Medical History
The characteristics of the 1286 participants and their medical history are outlined in Table 2. The majority of the surveys were completed by the primary caregiver of an individual with ASD (85%). More than half of the surveys were for children under 13 years old (54%), with 21% for teenagers and 16% for young adults (18 years or older). Seventy six percent of participants were male, and 24% were female. Autism was the most frequent diagnosis (43%), followed by Autism Spectrum Disorder (22%) and Asperger's syndrome (14%). The most common developmental history was "Abnormal development from early infancy, with no major regression or plateau in development" (32%). Furthermore, most participants received antibiotics during their first 3 years of life, with a median of 3 rounds. Most participants had moderate autism-related symptoms at age 3 years old (38%) and currently (38%).  1 . Grandparents were taken from those responded with "other" and noted they were grandparents. Numbers may not add up to 100% due to rounding.

Amino Acids
Amino Acids were rated as having a slight to moderate (1.1 to 1.6) overall perceived benefit with minimal AEs (0.1 to 0.4). For the amino acid blend, glutamine and taurine, the primary benefit was general benefit (43-57%) with small benefits in other symptoms. For tryptophan, the primary benefits were helping with falling asleep and staying asleep (see Table S2 and Figure 1).

Fatty Acids
Fatty Acids (FA) were rated as having a moderate to good benefit (1.2 to 2) with minimal overall AEs (0 to 0.2). For all FAs, the primary benefit was general benefit (32% to 59%), with secondary benefits in attention and cognition. See Table S3 and Figure 2.

Glutathione-Related Nutraceuticals
Glutathione-related nutraceuticals (including NAC) were rated as having a slight to moderate benefit (1.1 to 1.7) with minimal AEs (0 to 0.3). The most common benefit was general benefit (4% to 56%). See Table S3 and Figure 3.

Individual Minerals
Individual minerals were rated as having a slight to moderate benefit (1.3 to 2.1) with minimal AEs (0-0.3). The most common benefit was general benefit (15% to 70%). Lithium also helped with anxiety (24%), and magnesium helped with constipation (27%). Iron caused some gastrointestinal adverse effects in 17%. See Table S4 and Figure 4.

Multivitamins
Multivitamins were rated as having a slight to moderate benefit (1.4 to 1.9) with minimal AEs (0.0 to 0.2). The most common benefit was general benefit (50-55%). High dose multivitamin also improved general health (26%), and a high dose multivitamin, specifically designed for ASD, improved cognition (21%). See Table S7 and Figure 6.

Sleep-Related Nutraceuticals
Sleep-related nutraceuticals were rated as having slight to moderate benefit (1.2-2.1), with minimal AEs (0.1 to 0.3). The primary benefit was falling asleep (36-74%), followed by staying asleep (27-35%). For 5-HTP, there was also a general benefit (27%). See Table S8 and Figure 7. It is noteworthy that melatonin had the highest overall benefit score and was used by a very high number of participants.

Average of All Nutraceuticals
Averaging all the nutraceuticals reported in this paper, the average Overall Benefit and Overall AE was 1.6 (SD = 0.3) and 0.1 (SD = 0.1), respectively, reflecting that participants reported on average slight to moderate benefits with minimal adverse effects. Table 3 presents the top-rated nutraceuticals for 24 different symptoms. For most symptoms, nutraceuticals were moderately effective (net benefit scores >0.25), including aggression/agitation, anxiety, attention, cognition, constipation, diarrhea, general benefit, health, hyperactivity, irritability, language/communication, falling asleep, staying asleep, and social interaction/understanding. Other symptoms were only slightly affected (net benefit scores between 0.10 and 0.25) such as depression, eczema/skin problems, lethargy, obsessive-compulsive symptoms, reflux/vomiting, sensory sensitivity, stimming and tics/involuntary movements (Table 3).

Top Nutraceuticals by Symptom
It is important to note that less common problems, such as seizures, might receive lower scores since fewer individuals have these problems. These ratings should be interpreted cautiously, as they are averages, but they suggest which treatments families sensed were most helpful for a given symptom, which can potentially help guide treatment selection and future research.

Symptoms Nutraceuticals (Overall Net Benefit Rating)
General benefit, no one particular symptom

Overall Effects of Nutraceuticals
As a final part of this Survey, participants were asked to rate the overall effect of nutraceuticals (Table 4). A total of 77% of participants reported that nutraceuticals had a positive effect, ranging from slightly better (24%) to much better (27%), with 23% reporting no effect, and no reports that they generally resulted in worsened symptoms.

Analysis of the Effect on Specific Supplements on Change in Severity
To study the change in ASD severity related to nutraceuticals, nutraceuticals with at least 100 responses were selected in order to ensure there were enough cases to provide an adequate range of change in ASD severity. Since there were multiple categories of Omega 3 fatty acids (Fish Oils, Omega 3 Fatty Acids, Krill Oil) and B12 (oral and injected) these nutraceuticals were combined into categories. Thus, nutraceuticals selected included B12 (n = 170), Omega 3 fatty acids (n = 276), Epsom salt baths (n = 141), calcium (n = 110), magnesium (n = 153), zinc (n = 124), Vitamin C (n = 182), Vitamin D (n = 159), generic multivitamin (MVI) (n = 436), autism specific MVI (n = 103), and melatonin (n = 367). Because two other MVIs were being studied, high dose MVI (n = 45) was also included in the analysis. Because of the interest in the difference between injected vs. oral B12, the analysis was conducted on the separate groups of oral B12 (n = 127) and injected B12 (n = 76) as well as any B12 use. The analysis adjusted for baseline severity at age 3 years of age, developmental profile, number of rounds of antibiotic used in infancy, and gender.
The change in ASD severity was also related to baseline severity at 3 years of age in all of the analyses, which included taking B12 [χ(1) 2 = 336, p < 0.001], B12 injections In all models, a higher baseline severity was associated with a larger positive change in development as might be expected as higher severity patients have more potential for improvements. Almost all of the associations shown in Figure 9 demonstrate that treatment was associated with greater improvements. The exceptions were generic multi-vitamin, presumably because that meant participants did not take a multi-vitamin specific for ASD, and melatonin, probably because it treats a specific problem and is given to children with sleep disorders who may require additional non-nutraceutical treatments.  Figure 10B). Interestingly, the pattern of the child's development also affected the change in ASD severity when controlling for the benefit of injected B12 [χ(1) 2 = 24.32, p < 0.001]. This was due to the children with early onset ASD demonstrating significantly greater benefit (1.5) as compared to those who had a clinical regression and then a developmental plateau (−0.31), those with only a plateau (0.69) or those with a major developmental regression (0.46) when controlling for the perceived benefit of B12 injections.

Discussion
This study presents the Survey results of participants' reports of the perceived effectiveness and potential AEs of a wide range of nutraceuticals used in individuals with ASD. Nutraceuticals were generally reported to have a higher benefit compared to their AEs, with an average of 1.6 (slight/moderate benefit) and 0.1 (minimal AE), respectively. Reported benefits were generally in the slight/moderate range, and AEs were minimal.
Some of the nutraceuticals in this Survey have not been previously studied in ASD including an amino acid blend, glutamine, taurine, tryptophan, evening primrose oil, flax seed oil, krill oil, calcium, chromium, iodine, lithium, potassium, selenium, vitamin E, vitamin K, valerian root, Epsom salts, GABA, and milk thistle. Thus, this Survey provides preliminary data on the effects (both beneficial and adverse) of these unstudied treatments which can help guide researchers to choose the most promising treatments to study in the future.
Some of the treatments reviewed may not only help certain symptoms of ASD but also treat underlying metabolic abnormalities associated with ASD. For example, mitochondrial dysfunction is relatively common in individuals with ASD [10,63] and is potentially treated with carnitine, Coenzyme Q10, B vitamins, and vitamin C [64]. Oxidative stress is also commonly associated with ASD [13] and is potentially treatable with antioxidants such as folinic acid, MB12, vitamin C, vitamin E, glutathione, ribose, and NADH. Melatonin is also an antioxidant and has positive effects on mitochondrial function [65].
Furthermore, children with ASD have been found to have multiple abnormalities related to one-carbon metabolism, including lower plasma levels of methionine, S-adenosylhomocysteine (SAM), homocysteine, cystathionine, cysteine, and total glutathione (GSH), as well as significantly higher concentrations of S-adenosylhomocysteine (SAH), adenosine, and oxidized glutathione (GSSG) [7,8]. Some studies have demonstrated that many children with ASD have a partial blockage in the transportation of folates into the brain due to an autoantibody to the folate receptor alpha, the primary mechanism which transports folate across the blood-brain barrier [66,67]. High dose folinic acid (1-2 mg/kg/day) has been shown to be an effective treatment for children with ASD with primary improvements in language in a double-blind placebo-controlled study [24], consistent with the findings of this Survey. Also consistent with this Survey, an open-label study found that high-dose folinic acid is effective for improving attention in children with ASD who possess the folate receptor alpha antibody [66], and two other placebo-controlled studies have also reported improvements with folinic acid in ASD [25,26]. These positive studies on the benefits of folinic acid are consistent with the results of Table 3, which demonstrates that folinic acid and vitamin B12 are two of the top-rated treatments for many ASD-related symptoms.
These abnormalities in one-carbon metabolism often result in problems in methylation and transsulfuration in ASD, resulting in a reduction in the production of glutathione [68]. In fact, these abnormalities appear to be so prevalent that they may be diagnostic for ASD [69]. Several studies [70,71] have addressed treatment of these linked pathways by providing cobalamin and folate derivatives to supplement the linked methylation-folate pathway in order to enhance the production of glutathione, while other studies have supplemented glutathione directly [58]. The findings of these studies of the benefits of cobalamin, folate, and glutathione are consistent with the results of this Survey.
It is interesting to compare the results of this Survey for nutraceuticals versus the results of this Survey for pharmaceuticals reported previously [31]. Averaging all nutraceuticals and all pharmaceuticals, the nutraceuticals had significantly higher Overall Benefit (1.59 vs. 1.39, p = 0.01) and significantly lower Overall Adverse Effect (0.1 vs. 0.9, p < 0.0001), based on a 2-sided t-test of the medications that had 20 or more responses [31]. Caution is needed in interpreting these results, since there are substantial variations in ratings for individual treatments. However, in general, these findings suggest that nutraceuticals may be important treatment options for ASD, and more research into nutraceuticals and how they affect metabolism is warranted.

Strengths of This Study
One strength of this study is that some of these nutraceuticals have not been formally studied to date; therefore, this is the first data available on these treatments for ASD. Another advantage is that a uniform rating scale was used for all treatments, so that direct comparisons between different treatments could be made-this is often not possible for comparing data from clinical trials, since different assessment tools are typically used. Finally, another strength is the large number of participants in this study.

Limitations of This Study
There are several limitations of this Survey to consider. One limitation is that it is based on survey data, so there may be a significant placebo effect, especially since one of the most common benefits reported was "general benefit-no specific symptom". The ratings are based on perceived benefit (primarily by caregivers) and not by medical professionals. Age at which treatment was administered was not collected, which is a limitation of the study. Furthermore, there was no data collected on the dosages or durations of treatments (other than high versus low dose folinic acid). Therefore, various doses and durations of treatments may have been used by participants. Another limitation is the ASD-related diagnoses were not confirmed with standardized testing but were gathered by participant self-report. Finally, there is the potential for recall bias, where participants may not completely remember the effects of certain treatments. This may be reflected by the fact that no participants listed any of the nutraceuticals as causing worsening in ASD-related symptoms.

Conclusions
This Survey provides important information on the overall and specific benefits and adverse effects of 58 of the most commonly used nutraceuticals in ASD. The Overall Benefits were rated slightly higher for the nutraceuticals than for the most commonly used pharmaceuticals reported in the previous paper, with significantly lower ratings of adverse effects. The perception of participants of slight/moderate benefit with minimal adverse effects potentially explains why nutraceuticals were used by 75.2% of individuals with ASD in the Survey. This is consistent with the growing number of positive randomized clinical trials of nutraceuticals in ASD. Further research into nutraceutical treatments for treating biochemical differences and ASD symptoms is warranted.