What Is Known
- gastrointestinal disorders that may trigger behavioral symptoms;
- increased frequency of lactase deficiency;
- intestinal inflammation;
- increased intestinal permeability (“leaky gut”).
What Is New
- There was no difference in lactase, sucrase, maltase, and palatinase specific activity in autistic and nonautistic children who are evaluated for gastrointestinal disorders.
- The biomarkers calprotectin and lactoferrin identify children with and without autism who have intestinal inflammation, but are not more commonly found in children with autism.
- Abnormal intestinal permeability (lactulose/rhamnose ratio) in children with autism is not found more commonly in children with autism with gastrointestinal symptoms than in neurotypical children with similar symptoms.
Children with autism have behaviors that affect social interaction as well as verbal and nonverbal communication. Common problems such as gastroesophageal reflux or constipation may present with atypical symptoms such as stereotypical behaviors, aggression, or self-injurious behaviors. Consequently, gastrointestinal problems that may be easily recognized in a neurotypical child may go undiagnosed in a child with autism. Previous studies have suggested that lactase deficiency is common in both children with autism spectrum disorder (ASD) and neurotypical children (1–4). Some studies report that intestinal inflammation and increased intestinal permeability are more common in children with autism (5,6), whereas other studies could not find differences in intestinal permeability between autistic and neurotypical children (7). The aim of the present study was to assess intestinal function by determining intestinal permeability, mucosal inflammation, and disaccharidase activity in children with or without autism who were having a clinically indicated endoscopy.
Experimental Design and Patient Evaluation
The study population consisted of 61 children with ASD and 50 children with normal development. Children with autism were diagnosed by Diagnostic and Statistical Manual of Mental Disorders, 4th Edition determination by the treating physicians. Consent was obtained from at least 1 parent and children older than 7 years were asked to provide assent if possible. The study was approved by the institutional review board (Protocol No. 2004p-001279) and was carried out in accordance with the Helsinki Declaration of 1975. Information about the present study was registered into ClinicalTrials.gov. Children ranged in age from 18 months to 18 years inclusive. Before being approached to participate in the study all had clinical indications for diagnostic endoscopy and ileocolonoscopy determined by their gastrointestinal physician. The criteria used as indications for an upper endoscopy and/or colonoscopy in children with autism who had language were similar to the criteria used for neurotypical children, including diarrhea, blood in the stool, weight loss, regurgitation, food refusal, or abdominal pain (Table 1). Indications for children without language included chest tapping, head banging, posturing, aggressive/self-injurious behavior, or pushing on their abdomen. Subjects with bleeding disorders, or who had used pancreatic enzymes 14 days before the procedure were excluded from the study. We used a case-control design for the study. As was expected, the number of boys (41) in the autistic group was higher than the number of girls (20), whereas in the comparison group, the number of girls was higher (29 girls and 21 boys) (Table 1). The most common single gastrointestinal symptom in both groups was parental or patient report of abdominal pain followed by diarrhea. The frequency of abdominal pain was similar in both groups (38%), but diarrhea was reported more frequently in children with autism (36% vs 10%). A history of constipation along with abdominal pain was observed in 10% of children with autism but was not noted in controls.
All procedures were performed under general anesthesia. Esophageal, gastric, duodenal, ileal, and colonic biopsies were placed into formalin for routine pathological examination. Biopsies measured for disaccharidase-specific activity were obtained from the second part of the duodenum, snap-frozen in liquid nitrogen or dry ice in the endoscopy suite, and stored at −80o C until analysis. The carbohydrate solution used for the permeability test was introduced directly into the duodenum at the initiation of the endoscopy and urine was collected for the next 5 hours. Stool samples were collected by parents at home at least 2 days after the procedure, kept frozen, and delivered to the laboratory on ice.
Lactase, sucrase, maltase, and palatinase activity were performed on duodenal biopsies using the Dahlqvist (8) method and normalized to protein levels. Specific enzyme activity was expressed in μmol of hydrolyzed substrate min/g protein (U/g protein). Protein content was tested by the Bradford (9) method. Enzyme activity for lactase <15 U/g protein, sucrase <25 U/g protein, maltase <100 U/g protein, and palatinase <5 U/g protein was considered deficient (10,11).
The intestinal permeability test was performed after an overnight fast. A sugar solution (100 mL) containing 5 g lactulose and 1 g rhamnose was infused through the endoscope directly into the duodenum at the beginning of the upper endoscopy. Patients were not allowed any fluid orally for 2 hours after the solution was administered and food was withheld for the 5 hours during which time all urine was collected. The urine volume was measured and a 10-mL aliquot was preserved with 100 μL of 1% sodium thimerosal and stored at –20oC until analysis. Sugar analysis was performed using liquid chromatography with tandem mass spectrometry. A 50 μL aliquot from each urine sample is added to 200 μL of deionized water, and centrifuged at 12,000 g for 1 minute at ambient temperature. Fifty microliters, 0.5 mol/L aqueous sodium borohydride is added to 50 μL clear supernatant; after 30 minutes the reduction is quenched by addition of 50 μL, 0.5 mol/L aqueous acetic acid. The neutralized sample is desalted over solid phase extraction (Phenomenex Corp, Torrance, CA) (C18) and resolved in an Agilent 1200 high-performance liquid chromatography system: Stationary phase-150 × 2 mm, 5 μm aminopropylsilica; mobile phase: A—0.2% aqueous acetic acid; B—acetonitrile; 0.3 mL/min. The gradient is 7 minutes linear from 10% A to 25% A, 10 minutes isocratic 25% A, ramp down and equilibration with 10% A for 10 minutes. An Agilent 6400 mass spectrometer in negative ion mode detects rhamnose as the m/z = 165 peak and lactulose as the m/z = 343 peak, with standard (Sigma-Aldrich, Saint Louis, MO) calibration curves of peak areas linear from 0.05 to 10 μg/mL. Results were expressed as a percentage of lactulose and rhamnose recovery in urine after 5 hours excretion and as the lactulose/rhamnose ratio (L/R).
Fecal Biomarkers of Intestinal Inflammation
Stool samples were obtained at least 2 days after the procedure. An aliquot of 40 to 120 mg of stool was added to the extraction buffer. For calprotectin the weight/volume ratio was 1:50 and for lactoferrin the weight/volume ratio was 1:20. Each sample was vortexed for 30 seconds and then mixed on a shaker for 30 minutes. Fecal extracts were transferred to 1 mL tubes and centrifuged at 10,000 g for 20 minutes at 4°C. The supernatants were collected and frozen at −80o C. After thawing the extracts were diluted and run on enzyme-linked immunosorbent assay plates. Calprotectin was measured by the Calprest (PhiCal) enzyme-linked immunosorbent assay kit from Calpro (Norway); lactoferrin was measured using the BIOXYTECH Lacto f EIA kit (Northwest Life Science Specialties, Vancouver, WA).
Levels of calprotectin <50 μg of calprotectin/g stool are considered normal (http://www.calprotectintest.com/english/calprest.html). For healthy children and adults lactoferrin should not exceed 1.45 ± 0.40 μg/g (12).
Statistical analysis was performed using SPSS Statistics version 20 (IBM Corporation, NY). Pearson correlation was used to compare the different parameters between ASD and control group. Correlation coefficients were calculated for all possible combinations of pairs of variables. In addition, the data were further analyzed using independent samples t test. Nonparametric analysis was also performed using a nonparametric Mann-Whitney U test. Data are presented as mean ± standard error. A P value of <0.05 was considered statistically significant.
Morphological Analysis of Inflammation
Histologically, in the autistic group, 7 children out of 61 (11%) had esophagitis (Table 2). Five patients had features consistent with gastroesophageal reflux disease and 2 had features consistent with eosinophilic esophagitis. Six children (10%) had gastritis, 4 children (7%) had duodenitis, 3 children (5%) had ileal inflammation, and 12 children (19%) had colonic inflammation. Some children had inflammation in more than 1 area. Gastritis and duodenitis were nondiagnostic; neither Helicobacter pylori nor celiac disease were diagnosed. The findings in the ileum and colon also were non-diagnostic with increased cellularity in the lamina propria without chronic inflammatory changes. In the ASD group, 32 children out of 61 (52%) had some inflammation in the gastrointestinal tract, but it was generally mild and nondiagnostic. In children without autism, 4 (8%) had eosinophilic esophagitis, 2 had gastritis (4%), 2 had duodenitis (4%), 4 had ileal inflammation (8%), and 9 children (18%) had colonic inflammation. Three children had Crohn disease.
Specific activity for lactase, sucrase, maltase, and palatinase did not show any significant difference between autistic and nonautistic children (Table 3).
Fecal Biomarkers of Inflammation
Stool for calprotectin and lactoferrin analysis was measured in 49 children with autism and 36 children without autism. Samples from the 5 nonautistic children with Crohn disease or ulcerative colitis were analyzed as a separate group. Fecal calprotectin was elevated in 63% (31 out of 49) of children with ASD compared to 61% (19 out of 31) in nonautistic children. When the abnormal value was increased to >150 μg/g, fecal calprotectin was, however, elevated in 18% (9 out of 49) of children with ASD and in 26% (8 out of 31) of children without autism. Similarly, there was no difference in fecal lactoferrin between the 2 groups, as fecal lactoferrin was elevated in 10% (5 out of 49) of children with ASD and 6% (2 out of 31) in the non-ASD group (Table 4).
Intestinal Permeability Analysis
The urine for the intestinal permeability test was collected from 47 children with autism and 42 children with typical development. Lactulose (L) in the urine in children with autism was 1.4 times higher than that in nonautistic children, but urine rhamnose (R) was similar in both groups of children (Table 5). Lactulose/rhamnose ratio (L/R) in children with autism was 1.6 times higher than that in group of children with typical development; however, these differences between groups were not statistically significant.
Suspected gastrointestinal inflammation in children with autism has been the subject of much speculation and therapeutic intervention based on limited credible evidence. Many parents and clinicians have observed improvements in behavior, expressions of pain, and stooling patterns with dietary and nonevidence-based interventions. In data from the autism program at Cincinnati Children's Hospital, 24% of children with ASD had a history of at least 1 gastrointestinal symptom (13). Studies suggest that children who have more severe gastrointestinal symptoms as assessed by a Gastrointestinal Severity Index have more serious behaviors (14). The aim of the present study was to evaluate intestinal histology, inflammation, permeability, and digestion of disaccharides in a group of children with and without autism who were being evaluated for gastrointestinal symptoms. Children who were included in this study had clinical indications for an upper endoscopy or colonoscopy and children with autism who did not have stereotypical behaviors or gastrointestinal symptoms were not enrolled.
A “leaky gut” in children with ASD has been associated with behavioral changes (15). The permeability of the small intestine was assessed by measuring urinary excretion of orally administered sugars such as lactulose, L-rhamnose, and mannitol (M) is frequently used to characterize mucosal integrity and gut barrier function (16). Mannitol and rhamnose pass through the cell membrane (transcellular pathway) and characterize the absorptive function of intestinal mucosa, whereas lactulose passes through intercellular junctional complexes (paracellular pathway) and characterizes mucosal barrier function, the route taken by molecules with molecular mass >180 Da (17). A normal value for the L/R ratio is not easily identified from the literature. Miki et al (18) used >0.05, Van Elburg et al (19) (L/M) used >0.09, and Navarro et al (20) (L/M) used >0.1 as abnormal. Using 0.09 as the cut off there were 25 of 47 autistic children and 19 of 42 nonautistic children with abnormal L/R (P = 0.45), respectively. Similarly, choosing 0.1 as the cutoff, 22 of 47 autistic children had increased permeability compared to 12 of 42 nonautistic children (P = 0.08). Based on these data, intestinal permeability may be increased in some children with autism and those without autism. Furthermore, children with autism who have gastrointestinal disorders do not appear to be more likely to have a “leaky gut” when compared to neurotypical children with similar problems.
In children with autism, an increase in intestinal permeability was reported by D’Euphemia et al (21), who found an abnormally high lactulose/mannitol ratio in the urine in 43% of children with autism. De Magistris et al (5) found increased intestinal permeability in 37% of children with autism compared to 4.8% in healthy controls. Children who were on a gluten and casein-free diet had lower permeability. Interestingly, in these studies recovery of the transcellular permeability marker, mannitol, in the autistic group was similar to healthy controls, but recovery of the paracellular permeability marker, lactulose, was significantly higher than in controls who did not have intestinal disease or food allergy. In a group of 16 children with autism in Tunisia, 25% had increased intestinal permeability. The 4 children with abnormal permeability all had histologically normal appearing duodenal mucosa, but 2 had increased cellularity in the lamina propria. One had a positive immunoglobulin G endomysial antibody (22). Robertson et al (7) comparing 14 children with autism to healthy controls did not find a difference in intestinal permeability. A recent report by Navarro and colleagues (20), found that only 2 of 12 autistic patients (17%) had increased intestinal permeability and that gluten did not affect permeability. Variations in methodology and measurement of substrates may account for some of the variability in these results.
In the present study there was no difference in the permeability of the transcellular marker rhamnose. Although both the frequency of abnormal intestinal permeability as well as the lactulose level and the L/R ratio was higher in children with autism, the difference did not achieve statistical significance. Thus, our data do not support increased permeability in children with autism compared to neurotypical children undergoing an evaluation for gastrointestinal disorders. In the present study the intestinal permeability studies were done by instilling the solution directly into the duodenum while the patient was under general anesthesia. This technique avoids the slower emptying of the substrates when they are swallowed and released over a longer period of time from the stomach. After the solution was administered, duodenal biopsies were obtained and breaks in the mucosa could have affected absorption. The combination of medications used for anesthesia, the bolus of lactulose and rhamnose into the duodenum, and the duodenal biopsies may all contribute to the increase in intestinal permeability observed in both autistic and nonautistic subjects. These findings do not exclude the possibility that a subgroup of individuals with autism may have increased permeability.
Children with autism have been reported to have decreased disaccharidase activity, particularly lactase (1,2). Our previous study demonstrated that 58% of autistic children 5 years or younger and 65% of older children were lactase deficient (4). These findings are similar to the 66% frequency of lactase deficiency in children with ASD found in this study. Furthermore, there was no statistically significant difference in lactase or other disaccharidase activity between autistic children and nonautistic children with gastrointestinal symptoms.
Fecal calprotectin and lactoferrin are clinical biomarkers that are used to identify patients with colonic inflammation (23,24). In a group of 24 children with autism, there did not appear to be an increase in fecal calprotectin (25); however, De Magistris et al (5) found that approximately 25% of children with autism with or without gastrointestinal symptoms had elevated fecal calprotectin compared with approximately 12% of normal relatives. In patients with Crohn disease values of fecal calprotectin <170 μg/g, however, correlate with remission of disease (26). The mean fecal calprotectin level in the children with autism in this cohort was 111 μg/g well below the cut off for what is considered to be active inflammation in Crohn disease. These low levels of calprotectin and lactoferrin are most consistent with insignificant inflammation.
Clinicians struggle with the value of initiating a diagnostic evaluation of gastrointestinal function in a child with autism who may have atypical symptoms. Common disorders such as slow transit constipation may result in colonic distension causing pain or dysbiosis that can be associated with increased gas or diarrhea. The role of dietary restriction remains a confounding factor. The present study suggests that the frequency of inflammation, intestinal permeability, or brush border enzyme activity does not differ in children with or without autism who are being evaluated for intestinal disorders. The technical issues involved in performing the intestinal permeability test may, however, have affected the results in both groups. The data from the present study should not be used to justify performing endoscopic evaluations in all children with autism. Only children suspected of having gastrointestinal disorders were included. The results of the present study suggest that common gastrointestinal problems occur in children with autism and should be evaluated. There is no evidence to support that gastrointestinal disorders cause autism. Identifying children with ASD who have concomitant medical conditions such as inflammatory bowel disease, malabsorption, or lactose intolerance may be challenging because their symptoms are atypical. In these children, criteria such as stereotypical or self-injurious behaviors may identify those individuals worthy of additional evaluation.
This research was sponsored in part by Cure Autism Now and Autism Speaks, Inc. Additional funding for this project came from philanthropic support to Mass General Hospital for Children from Martin Schlaff and James Brooks. This manuscript has been read and approved by all authors. This article is unique and not under consideration by any other publication and has not been published elsewhere. The statements, findings, conclusions, and recommendation are those of the author(s) and do not necessarily reflect the view of Cure Autism Now, Autism Speaks, Inc, or other sponsors.
1. Horvath K, Papadimitriou JC, Rabsztyn A, et al. Gastrointestinal abnormalities in children with autistic disorder. J Pediatr
2. Horvath K, Perman JA. Autism and gastrointestinal symptoms. Curr Gastroenterol Rep
3. Ashwood P, Anthony A, Torrente F, et al. Spontaneous mucosal lymphocyte cytokine profiles in children with autism and gastrointestinal symptoms: mucosal immune activation and reduced counter regulatory interleukin-10. J Clin Immunol
4. Kushak R, Lauwers GY, Winter HS, et al. Intestinal disaccharidases activity in patients with autism: effect of age, gender, and intestinal inflammation. Autism
5. De Magistris L, Familiari V, Pascotto A, et al. Alterations of the intestinal barrier in patients with autism spectrum disorders and in their first degree relatives. J Pediatr Gastroenterol Nutr
6. Souza NC, Mendonca JN, Portari GV, et al. Intestinal permeability
and nutritional status in developmental disorders. Altern Ther Health Med
7. Robertson MA, Sigalet DL, Holst JJ, et al. Intestinal permeability
and glucagon-like peptide-2 in children with autism: a controlled pilot study. J Autism Dev Disord
8. Dahlqvist A. Assay of intestinal disaccharidases. Anal Biochem
9. Bradford MM. A refined and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem
10. Pfeffercorn MD, Fitzgerald JF, Croffie JM, et al. Lactase deficiency: Not more common in pediatric patients with inflammatory bowel disease than in patients with chronic abdominal pain. J Pediatr Gastroenterol Nutr
11. Belmont JW, Reid B, Taylor W, et al. Congenital sucrase-isomaltase deficiency presenting with failure to thrive, hypercalcemia, and nephrocalcinosis. BMC Pediatr
12. Kane SV, Sandborn WJ, Rufo PA, et al. Fecal lactoferrin is a sensitive and specific marker in identifying intestinal inflammation. Am J Gastroenterol
13. Molloy CA, Manning-Courtney P. Prevalence of chronic gastrointestinal symptoms in children with autism and autistic spectrum disorders. Autism
14. Adams JB, Johansen LJ, Powell LD, et al. Gastrointestinal flora and gastrointestinal status in children with autism—comparisons to typical children and correlation with autism severity. BMC Gastroenterol
15. De Magistris L, Picardi A, Sapone A, et al. Comprehensive Guide to Autism (Intestinal Barrier in Autism). 2014; New York: Springer, 2047-2060.
16. Bjarnason I, MacPherson A, Hollander D. Intestinal permeability
: an overview. Gastroenterology
17. Travis S, Menzies I. Intestinal permeability
: functional assessment and significance. Clin Sci
18. Miki K, Butler R, Moore D, et al. Rapid and simultaneous quantification of rhamnose, mannitol, and lactulose in urine by HPLC for estimating intestinal permeability
in pediatric practice. Clin Chem
19. Van Elburg RM, Uil JJ, Kokke FT, et al. Repeatability of the sugar-absorption test, using lactulose and mannitol, for measuring intestinal permeability
for sugars. J Pediatr Gastroenterol Nutr
20. Navarro F, Pearson DA, Fatheree N, et al. Are ‘leaky gut’ and behavior associated with gluten and dairy containing diet in children with autism spectrum disorders? Nutr Neurosci
21. D’Euphemia P, Celli M, Finocchiaro R, et al. Abnormal intestinal permeability
in children with autism. Acta Pediatr
22. Boukthir S, Matoussi N, Belhadj A, et al. Abnormal intestinal permeability
in children with autism. Tunis Med
23. Tibble JA, Bjarnason I. Fecal calprotectin as an index of intestinal inflammation. Drugs Today
24. Tibble JA, Bjarnason I. Non-invasive investigation of inflammatory bowel disease. World J Gastroenterol
25. Fernell E, Fagerberg UL, Hellström PM. No evidence for a clear link between active intestinal inflammation and autism based on analyses of faecal calprotectin and rectal nitric oxide. Acta Paediatr
26. Cerrillo E, Beltrán B, Pous S, et al. Fecal calprotectin in ileal Crohn's disease: relationship with magnetic resonance enterography and a pathology score. Inflamm Bowel Dis