This study was approved by the appropriate Institutional Review Board in accordance with local regulations and informed consent/assent was obtained from the subjects and/or their legal guardians after the nature of the procedures had been explained.

Male and female children (n = 5) aged 9-14 years, diagnosed with Autism Spectrum Disorder (ASD) and meeting the inclusion criteria below, were recruited sequentially through an out-patient clinic. If the caregiver expressed interest in their child participating in the study, the consent form was reviewed with the family and informed consent and assent (if applicable) obtained. Inclusion criteria were as follows: 2-21 year-old males and females with ASD with a diagnosis of inflammatory bowel disease. The participant had to be under the care of a caregiver.

Willing to participate in the study by attending regularly scheduled appointments and completing the necessary measures, and be accessible to a research staff member via phone or in-office on a weekly basis throughout the study intervention period. Exclusion criteria were as follows: pregnant or nursing females up to 21 years of age; requirement of receiving any nutrition via enteral or parenteral routes; infectious colitis, metabolic disorders such as phenylketonuria, leukodystrophy, lysosomal disorder, Wilson’s disease, and other concurrent physical or mental disorders that preclude participation in assessment procedures. Concurrent medications, dietary restrictions, and/or therapies were not exclusions to participation. However, concurrent interventions could not change during the period of the trial, unless there was a clear medical indication to do so.

Enrolled participants received total elemental nutrition for 12 weeks under the guidance of a registered dietitian and a clinician. Anthropometric measurements were collected at baseline, 4, 8, and 12 weeks, and at two follow-up appointments conducted at 6 and 12 months. Biochemical markers and dietary analysis were monitored throughout the study, and caregivers participated in weekly phone or in-office visits to address any concerns raised by the clinician, dietitian and/or caregiver.

The study procedures performed during each phase are summarized in Table 1.

Caregivers were given standardized written instructions and measurement guidelines to record all food and beverage intake for two weekdays and one weekend day that they considered typical for their child (Three-Day Food Diary, Supplementary Figure S1). Homemade recipes, brand names of packaged products and restaurant information of foods consumed were required. Food records were analyzed using Food Processor SQL (version 10.3.0, 2008, ESHA, Salem, OR; Supplementary Figure S2). Supplement data was not included in the nutrient analysis for an accurate assessment of dietary nutriture. Ideal body weight (IBW) was calculated as the appropriate weight for height according to the CDC growth charts. Estimated energy requirement for each patient was derived from height, weight, and a moderate physical activity level using the formulas developed by the Institute of Medicine 41.

A clinical evaluation was performed at baseline and at the completion of the study to assess the health and well-being of participants. Participants were also monitored throughout the study to assess actual elemental nutrition intake and tolerance, and to ensure that caloric intake was adequate, hydration status normal, and nutritional goals were achieved and maintained. Caregivers were given a standard weight scale and received training on proper use in order to track weight outside of the clinic. Greater than 5% weight loss in one month was considered abnormal, and would either result in the clinician either changing the dosing of the elemental formula or withdrawing the participant from the study. Gastrointestinal tolerance was also assessed weekly using the Gastrointestinal Symptoms Rating Scale (GSRS) 42, which document issues with diarrhea, indigestion, constipation, abdominal pain, and GERD on a scale of 0-3. Severe side effects, such as an allergic reaction, severe abdominal pain, vomiting, aggression and/or hyperactivity interfering with daily functioning, or significant alterations in white blood cell count, or liver or pancreatic enzymes, would have resulted in the elemental formula being withdrawn.

Total daily nutrition was provided by elemental formulas that were provided at no cost to study participants. Three options for elemental nutrition were offered to the participants and their caregivers: two flavored powdered elemental formulas, Ross Nutricia Neocate and Elecare, and the ready-to-drink elemental formula, Neocate EO28 Splash. The powered formulas were prepared to provide appropriate caloric density to meet estimated energy requirements for weight maintenance or weight gain if indicated. Neocate EO28 Splash required no additional preparation. Two participants favored the ready-to-drink formula, while the other three chose the powdered formula. Formula was prescribed by the clinician and caregivers received sufficient formula for the 3 months study intervention period. Verbal and written instructions on titration and maintenance dose administration was provided to the caregiver under the guidance of a registered dietitian. Each participant received the optimal amount of formula appropriate for their daily energy and nutrient needs. Feedings occurred at meal or snack time, or ad libitum.

A daily food (formula) diary was maintained by the caregiver to record: i) all feedings with the elemental formula consumed on a daily basis, including documentation of any missed feedings and reasons for feedings being missed; ii) any ingested foods and beverages that were not allowed in the dietary intervention; and iii) suspected adverse diet-related events including the nature, severity and duration of such events. The daily food diaries were discussed during the weekly study appointments. At the end of 12 weeks, the clinician and dietitian reviewed all the study data and subsequently recommended that either the participant continue with elemental nutrition (outside of the study) or they begin the process to reintroduce solid foods, done under the guidance of a registered dietitian for a period of 4 weeks. In all cases, whether ED was continued or discontinued, follow-up assessments at 6 and 12-months were scheduled although not all families provided additional data past the study intervention period.

Primary outcomes examined the efficacy of the elemental formula, as evidenced by the improvement in bowel symptoms and height/weight gain. Secondary outcomes included assessments of nutritional status and nutrient intake based on measurements of blood concentrations of total protein, albumin, pre-albumin, ferritin, C-reactive protein, RBC, MCV, MCHC, AST and alkaline phosphatase. Vitamin and minerals that were assessed included: sodium, potassium, chloride, magnesium, calcium, phosphorous, zinc, vitamin D, E, and K and folate. Nutrient intake was assessed by comparing submitted 3-Day food diaries with baseline data.

A 14-year old boy with Crohn’s disease presented with abdominal pain, frequent daily loose stools, and low muscle tone. BMI was <1^st percentile. Following the 12-week intervention, the participant reported a reduction in abdominal pain and improvement in stool consistency and frequency. Although there was a small overall loss in body weight from 94.5 lbs to 89 lbs from baseline to 12 weeks, this subject reported a 34% weight gain at the 6-month follow-up appointment (Table 2). Height remained unchanged during this time. BMI remained in the <1st percentile from baseline to 12 weeks but increased to 23^rd percentile at 6 months. Intestinal permeability (L/M recovery test) remained within normal limits throughout the study. CRP decreased from a baseline value of 4.4 mg/dL (high) to <0.5 mg/dL (reference range <0.8mg/ml) and was within normal limits (WNL) at 12 weeks, although it increased to 1.7 and 3.0 mg/dL at the 6- and 12-month follow-up appointments, respectively. Vitamin D status increased from a baseline level of 17 ng/mL (insufficient reference range: <20ng/ml) to 61 ng/mL (optimal reference range: 30-100ng/ml) at 12 weeks, but decreased to 28 and 20 ng/mL (suboptimal reference range: 20-29 ng/ml) at the 6- and 12-month follow-up appointments, respectively. All other blood work remained WNT throughout the study. Nutrient intake was reported as 2631 total kCal, 362g total carbohydrates, 82g total fat and 36g total protein at baseline. After the 12-week study intervention period, nutrient intake was reported as 2000 total kCal, 210g total carbohydrates, 95g total fat and 60g total protein. GSRS analysis reported a total score of 13 at baseline and 1 at 12 weeks. This was attributed to a decrease in the pain and diarrhea domains after the 12-week study intervention. The caregiver also reported that the participant had no abdominal pain and “stool were daily and formed”. Although there were no formal behavioral assessment of the participants who were included in the study, a parental caregiver reported anecdotes of increased levels of physical activity, and improved socializing with peers. Solid food was re-introduced at 12 weeks as per recommendations but no nutrient intake data was provided at follow-up appointments. Following the re-introduction of solid food, the participant’s weight increased to 120lbs by the 6-month appointment and remained unchanged at the 12-month appointment.

An 11-year old boy diagnosed with gastropathy and enteritis and a history of pica presented with chronic constipation cycling with diarrhea, consistent abdominal pain prior to defecating, and chronic eructation. Weight, height and BMI were considered to be in a healthy range at baseline. After the 12-week study intervention, caregivers and/or participants reported improvements in GI symptoms and reduction in pain. The participant showed a consistent increase in body weight from 85.2 lbs at baseline to 93 lbs at 12 weeks, representing an increase of 7.8 lbs (Table 3). At the 6- and 12-month follow-up appointments, his weight was 88.25 and 92 lbs, respectively. Height increased slightly during the intervention period from 59 to 61.5 in but had not increased further at follow-up. BMI remained relatively steady throughout the study (Table 2). Lab values assessing nutrition and inflammation remained WNL except for vitamin D, which was considered suboptimal at 29 ng/mL at the 12-month follow-up appointment. Nutrient intake was reported as 2489 total kCal, 251g total carbohydrates, 126g total fat and 98g total protein at baseline. After the 12-week study intervention period, nutrient intake was reported as 2060 total kCal, 204g total carbohydrates, 92g total fat and 58g total protein. The participant’s mother reported some dietary infractions during the study, as well as periods of hyperactivity, visual stimming, bedwetting, and daytime incontinence. Pica (with items such as rocks, playground bark, and leaves) was also reported during the first month of intervention. The participant was subsequently diagnosed with round worms and began a parasite protocol under the care of his PCP. Treatment was completed half way through the ED intervention period. Follow-up nutrient analysis demonstrated an increase in total calories, with macronutrients reported as similar to baseline levels (data not shown). GSRS analysis reported a small improvement in GI symptoms with a score of 6 at baseline and 4 at 12 weeks. While the diarrhea domain score had decreased, the constipation domain score had increased. Anecdotal reports from a parental caregiver described a reduction of abdominal pain after the intervention period, but bed-wetting was still a frequent problem. Solid food was re-introduced at 12 weeks as recommended. Nutrient intake data provided at the 6- and 12-month follow-up appointments indicated that the participant was receiving appropriate nutrition and weight was stable over this time.

A 13-year old boy with a history of esophageal ulcers and lymphonodular hyperplasia of the terminal ileum presented with abdominal pain and diarrhea. Weight, height and BMI were considered to be in a healthy range at baseline. After the 12-week study intervention, the participant’s weight increased from 104 lbs at baseline to 115.4 lbs at 12 weeks representing an increase of ~11 lbs during the intervention period (Table 3). At the 6- and 12-month follow-up visits, his weight had increased to 118 and 119 lbs, respectively. Height increased by 4 in during the intervention period (65.5 to 69.5 in) remaining the same at both follow-up visits (Table 4). BMI remained relatively stable, decreasing slightly at 12 weeks due to the gain in height. Intestinal permeability decreased from 0.18 (reference range: <0.1) to 0.05 after the 12-week intervention and remained WNL at follow-up. Hemoglobin and hematocrit values were slightly high at baseline but within range at all other timepoint measured. (data not shown) All other lab work remained WNL. Nutrient intake was reported as 2517 total kCal, 259g total carbohydrates, 108g total fat and 128g total protein at baseline. After the 12-week study intervention period, nutrient intake was reported as 2844 total kCal, 415g total carbohydrates, 100g total fat and 71g total protein. GSRS analysis did not highlight any specific GI symptoms at baseline or at 12 weeks. Solid food was re-introduced at 12 weeks as recommended. Nutrient intake data provided at the 6- and 12-month follow-up appointments indicated that the participant was receiving appropriate nutrition and weight had increased by 4lbs at the 6-month time point, remaining stable at the 12-month follow up.

A 9-year old boy, significantly underweight and diagnosed with gastropathy, esophageal ulcers, and aphthous ulceration of the rectosigmoid presented with abdominal distention and diarrhea. The participant’s weight increased from 49.5 lbs at baseline to 52.8 lbs at the end of the 12-week intervention period representing an increase of ~3.5 lbs. Weight gain continued after transitioning back to solid food with a reported weight of 54.4 and 67 lbs at the 6- and 12-month follow-up appointments, respectively placing him at the 3.6th and 20th percentile, respectively (Table 5). Data for height, and therefore BMI, during the study was not available. All lab measures remained WNL throughout the study. Nutrient intake was reported as 1992 total kCal, 210g total carbohydrates, 91g total fat and 60g total protein at baseline. At the end of the 12-week study intervention period, nutrient intake was reported as 1659 total kCal, 242g total carbohydrates, 60g total fat and 42g total protein. GSRS analysis indicated a baseline score of 11 that was attributed to all 5 GI domains assessed. Unfortunately, a GSRS was not completed at 12 weeks, although data was available at 8 weeks, which indicated that GI symptoms were resolving: a score of 6 was reported primarily attributed to decreases in the pain and indigestion domains. Solid food was re-introduced at 12 weeks with no concerns. Nutrient intake data provided at the 6- and 12-month follow-up appointments indicated that the participant was receiving appropriate nutrition. Clinical follow-up to assess the status of previous gastrointestinal pathology after the study was not reported to study investigators.

A 9-year old boy with a history of enuresis, GERD, and lymphonodular hyperplasia of the terminal ileum and colon presented with abdominal pain, diarrhea, and vomiting. Weight, height and BMI were considered to be in a healthy range at baseline. There was no significant weight change over the 12-week intervention period however the child grew 1.25 inches (Table 6). Weight and height continued to increase at both the 6- and 12-month follow-up appointments with a final weight of 74 lbs and height of 57.5 in reported at the 12-months follow-up appointment. Lab values remained WNL throughout the study except for intestinal permeability, which decreased from 0.11 (reference range <0.10) to 0.06 after the 12-week intervention, and remained WNL at 6 months (a 12-month follow was not available). Nutrient intake was reported as 2662 total kCal, 299g total carbohydrates, 131g total fat and 93g total protein at baseline. After the 12-week study intervention period, nutrient intake was reported as 2133 total kCal, 312g total carbohydrates, 75g total fat and 54g total protein. GSRS analysis reported an improvement in stool consistency with a score of 22 at baseline and 8 at 8 weeks (a 12-week GSRS was not available). This was primarily attributed to decreases in the pain, reflux and indigestion domains. Parental caregiver reported initial concerns over pica (paper) when ED was started so gave ice-chips to provide a chewing sensation, which helped although the pica did not completely resolve during the study. Improved activity levels towards the end of the intervention period was reported, although enuresis did not resolve. Solid food was re-introduced at 12 weeks as recommended with no concerns reported. Nutrient intake data provided at the 6- and 12-month follow-up appointments indicated that the participant was receiving appropriate nutrition and their weight had increased by 4lbs.

This study was funded by The Jane Botsford Johnson Foundation. We are very grateful to all the families that participated in this research study. We would also like to thank the staff at The Johnson Center for implementation of this study protocol. This study was approved by the Austin Multi-Institutional Review Board (AMIRB).

This study was approved by the appropriate Institutional Review Board in accordance with local regulations.