November 2009  << Back  

  Can't open the PDF? Click here for help.



The possibility and probability of a gut-to-brain connection in autism

K. L. Reichelt, MD, PhD

Department of Pediatric Research, Rikshospitalet Medical Centre, University of Oslo, Oslo, Norway

A. M. Knivsberg, PhD

The Reading Centre, University of Stavanger, Stavanger, Norway

BACKGROUND: We have shown that urine peptide increase is found in autism, and that some of these peptides have a dietary origin. To be explanatory for the disease process, a dietary effect on the brain must be shown to be possible and probable.

METHODS: Diagnosis was based on DSM-III and DSM-IV criteria. We ran first morning urine samples equivalent to 250 nm creatinine on high-performance liquid chromatography (HPLC) reversed phase C18 columns using trifluoroacetic acid acetonitrile gradients. The elution patterns were registered using 215 nm absorption for largely peptide bonds, 280 nm for aromatic groups, and 325 nm for indolyl components. We refered to a series of published ability tests, including Raven’s Progressive Matrices and the Illinois Test of Psycholinguistic Ability, which were administered before and after dietary intervention. The literature was also reviewed to find evidence of a gut-to-brain connection.

RESULTS: In autistic syndromes, we can show marked increases in UV 215-absorbing material eluting after hippuric acid that are mostly peptides. We also show highly significant decreases after introducing a gluten- and casein-free diet with a duration of more than 1 year. We refer to previously published studies showing improvement in children on this diet who were followed for 4 years and a pairwise matched, randomly assigned study with highly significant changes. The literature shows abundant data pointing to the importance of a gut-to-brain connection.

CONCLUSIONS: An effect of diet on excreted compounds and behavior has been found. A gut-to-brain axis is both possible and probable.

KEYWORDS: autism, peptides, proteins, uptake, diet



Autism has a solid genetic disposition.1,2 This necessarily entails chemical changes. However, genetic disorders that become evident over time probably depend on environmentally supplied compounds that lead to accumulation of metabolites that are proximal to the genetic deviation and decrease distal to the lesion. This is like a dam in a river, which causes differently shaped lakes to be formed, depending on topography, and secondary pathways (overflow) to open when full. In humans, our constitution is the “topography,” so that one genetic defect may lead to different phenotypes. Toxicologic inhibition of, or lack of, cofactors for enzymes in the same pathway may lead to similar results and may mimic or reinforce a purely genetic state of affairs.

The FIGURE represents a metabolic chain in which E1, E2, etc, are different enzymes and p is a negative feedback inhibitor on the rate-limiting enzyme, E1. With insufficiency of enzyme following D indicated by a vertical block, D will increase; since feedback is prevented by a decrease in e, f, g, and especially p, the system can easily run wild. Removing input a will reduce D, and supplying cofactors to the enzyme following D may increase its efficiency.3 Cofactors are usually derived from vitamins, trace minerals, and—if membrane bound—by increasing conformational flexibility by a more fluid membrane (unsaturated fatty acids). An excellent example of the simplistic outline shown above is phenylketonuria (Föllings disease), in which an inability to metabolize the amino acid phenylalanine (F) causes accumulation of this amino acid and opening of secondary pathways. Reducing the input of F prevents severe mental retardation.

FIGURE Consequences of an enzyme insufficiency to a metabolic chain

Conditions that must be fulfilled

To be credible, the idea that dietary peptides and/or proteins influence the behavior and development of autistic children, the following conditions appear necessary:

  1. It must be physiologically possible and probable.

  2. There must be demonstrable chemical changes that support this view.

  3. It must be shown with reasonable probability that removing the offending source of the accumulated compounds is effective and possible, in order to decrease the input.

  4. The pathophysiology of the disease ought to be plausibly explained.

  5. Supportive evidence from other disorders that may have similar pathophysiology should exist.

1. Physiologically possible—and probable

If low molecular weight peptides from food proteins are the mediators of the autistic state, peptides should be formed in the gut or systemically from uptake of precursors. Peptides are formed in the gut in animals4 and humans.5 If proteins are taken up and the breakdown is insufficient, peptides will accumulate. Intact uptake of protein and peptides is found in animals and humans,5,6 and peptides taken up from the gut can be found in urine.6 Decreased breakdown of peptides in the intestinal mucosa increases the uptake of specific peptides in animals.7,8 Exorphins are especially peptidase resistant because of their structure.

Several investigations have shown uptake of intact protein and enzymes in humans9 and animals.10 This uptake can be measured as increased IgA antibodies to such proteins.11 That proteins are taken up in normal persons is strongly supported by the presence of proteins in the breast milk of mothers who ingested these proteins.12-15 Also botulinum toxin, a protease, is taken up, passes the blood–nervous system barrier, and cuts one peptide bond in SNAP-25 protein in the synapses, with fatal results.16 The prion, a protein introduced through diet, retains its tridimensional structure, acting like a chaperone toward an endogenous protein, which causes misfolding and spongiform encephalitis. This example may illustrate why active enzymes can also pass into the bloodstream.

Inflammation of the mucosa reduces the activity of lytic enzymes such as those seen in celiac disease, resulting in increased uptake of protein and peptides.17 In autism, inflammatory cytokines are apparently released by milk proteins and gluten/gliadin.18 Several groups have reported panenteritis in autism, which would result in increased uptake from the gut due to mucosal damage and loss of membrane-attached enzymes.19,21 Inflammatory cytokines can alter the epithelial barrier to food antigens in disease.22

Exorphins are able to penetrate the blood-brain barrier in animals23 and humans.24 The results of peripheral (IV) application of casomorphin 1-7 on behavior in rats25 as well as in inducing immediate early antigen Fos26 blocked by naloxone, reinforce this view. Gluten exorphin B5 also causes hyperprolactemia in rats27 after peripheral administration. The behavioral effects in rats were the same as those obtained by intracranioventricular injections,28 and the opioids were found to cause inhibition of dopamine uptake and dopaminergic hyperactivity.28,29

Because the basic mechanisms are in place, we may therefore conclude that peptides and proteins are taken up from the gut and can produce bioactive peptides. Also, because the blood-brain barrier is peptide permeable, CNS effects are to be expected.

2. Demonstrable chemical changes

Increased secretion of peptides in urine has been found in patients with autism and schizophrenia.30-34 TABLE 1 shows the total amount of UV 215-absorbing material eluted after hippuric acid on high-performance liquid chromatography (HPLC) reversed-phase columns; 215 nm absorption is taken as a measure of peptide-like bonds (UV 280 nm absorption is measured simultaneously for aromatic groups). The difference from controls is strongly significant. In TABLE 1, the P values were obtained by comparing each country with controls in the same age range. Urine equivalent to 250 nm creati-nine from the first morning urine was run on HPLC C18 reversed-phase columns in a trifluoroacetic acid (TFA, 10 mM) acetonitrile gradient.34 Two-tailed, unpaired t test with Welch correction was used. The very large SDs (standard deviations) are mainly due to some excessively high levels in patients with regressive autism.

In addition, IgA antibodies increased against gluten, gliadin, and casein but not lactoglobulin, lactalbumin, and ovalbumin.30,32,35 This means that, without celiac disease, uptake of protein or its epitopes across the mucosa is increased.

Many studies show intestinal and mucosal damage in autistic children.19,20,36,37 This could explain the increased gut permeability shown.38 Exposing children on celiac diet treatment to gliadin/gluten during provocation caused long-lasting EEG changes.39 Nuclear magnetic resonance (NMR) spectroscopy of the brain after intake of food in chronic ileitis in humans shows perivascular edema in the white substance shortly after food intake.40,41

Increased urine peptide levels reflect decreased breakdown due to solely genetic, genetic and toxicologic, or solely toxicologic reasons. When the normal breakdown is deficient, peptides are regularly found in the urine of animals42-44 and humans.45 If the peptide level is below the reuptake capacity of the kidney tubules, near-normal peptide levels in the urine may be expected, as seen in 80% of high-functioning autistic children.46


Total level of UV 215 nm–absorbing material eluted after hippuric acid—a comparison of different countries vs controls

Scandinavia Italy Serbia Slovenia Turkey Controls
Average 620 403 584 564 427 289
SD 471 307 504 523 273 84
No. 315 145 139 51 26 181
95% CI 560 to 773 343 to 453 500 to 667 416 to 711 317 to 537 277 to 302
P value <.0001 <.0001 <.0001 .0005 .017
CI: confidence interval.
3. Effect of diet

Since the exorphins are derived from casein, gliadin, and gluten, and IgA antibodies against these proteins are statistically increased in children with autism, a reasonable course was to try to remove these proteins. This is similar to removing phenylalanine in phenylketonuria (Föllings disease). Most interventions carried out for a sufficient period of time (months to 1 year) register positive effects.30,32,47,51 The open study47 was followed for 4 years49 to counteract placebo effects (TABLE 2). A paired, single-blind, randomly assigned study confirmed the earlier data (TABLE 3).51 Over the Internet, a great number of before and after registrations have been done52 and show an amazing variation in the rate of recovery. Data from the Autism Institute in San Diego, California, point to the same.53 Negative reports are from trials conducted for very short time periods, with extremely varying age and no control of compliance. After puberty, the response to diet is very mixed.54 It is possible that enzyme supplementation may replace diet in the future.55 Many case reports have also been positive.56

Since vitamins and trace minerals are cofactors for many enzymes, it makes sense that they may influence the direction of reactions by mass action and also conformational change of the Km for substrates.3 However, very high levels of vitamin treatment and trace minerals over long periods of time are without published peer-reviewed data, except for B6 and magnesium.57 Other effects may also be present, such as the mercury-chelating effect of vitamin C,58 resulting in improvement of autistic symptoms.

The total time for each patient on a diet free of gluten and casein is 2 years. UV 215 units were as described in TABLE 1. Wilcoxon matched-pairs signed rank test 2-tailed was used. Pairing was effective. The decrease in peptide levels was found despite abundant intake of meat and fish (TABLE 4), further supporting the view that the peptide compounds do come from casein and gluten. Similar changes were found in the open study. No changes other than diet were made in the experimental group, which was given a gluten- and casein-free diet for 1 year. Mann-Whitney U test 2-tailed was used. (For details and tests used, see the original paper.51)


Measured behavioral changes from the first open test—followed for 4 years to reduce placebo effects47,49

Test Initial score ± SD 1-year change ± SD 4-year change ± SD No. P value
C-Raven 6.8 ± 2.8 +8.6 ± 2.8 +8.6 ± 3.2 12 .005*
ITPA 25.7 ± 5.5 +2.7 ± 2.5 +6.1 ± 2.8 10 .005*
Social isolation 8.5 ± 3.3 -6.1 ± 2.7   14 .005
Bizarre traits 5.3 ± 2.2 -5.3 ± 1.22   14 .005
DIPAB: Diagnose af Psykotisk Atfærd hos Børn (diagnosis of psychotic behavior in children); ITPA: Illinois Test of Psycholinguistic Abilities.
*Denotes the 4-year probability score. Mann Whitney U tests were used for all comparisons. Note that Raven Progressive Matrices reach a maximum after 1 year, whereas ITPA did so only after 4 years. This may reflect the greater demand for an intact brain to learn language.


Single blind pair-wise randomly assigned autistic children and the 1-year effect of gluten- and casein-free diet compared to autistic children without this intervention51

  Before After Diet to control, significance, P value No. Within-group significance, P value
Total score for autistic symptoms          
Diet group 12.5 ± 2.2 5.6 ± 2.4 10 .005
Control group 11.5 ± 3.9 11.2 ± 5.0 .001 10 .798
Subscore for social isolation, aloofness          
Diet group 7.6 ± 1.7 3.0 ± 1.4   10 .005
Control group 7.1 ± 2.8 6.2 ± 2.9 .003 10 .337
Subscore for communicative skills, social contact          
Diet group 3.9 ± 0.9 6.2 ± 1.1   10 .007
Control group 4.3 ± 1.3 4.5 ± 1.6 .004 10 .623
Subscore for strange and unusual behavior          
Diet group 4.9 ± 1.5 2.6 ± 1.7   10 .007
Control group 4.5 ± 1.6 4.8 ± 2.6 .007 10 .518


Peptide levels before and after dietary intervention in the same patients

  Total before Total after Autism peak before Autism peak after
Mean 425 204 48 26
SD 299 112 24 13
No. 108 108 107 107
P value <.0001 <.0002
4. Explaining the pathophysiology

We ought to be able to explain many physiological aspects of autism:

  • increased brain size in the first 2 years of life and later reductions in volume and development59

  • lack of social empathy and typical aloofness

  • stereotopies and rituals

  • poor habituation

  • emotional, uncontrolled outbursts

  • hyperacuity

  • immunologic changes

  • genetic disposition and recent increases in the number of children with autism.

Increase in peptides would lead to a general peptidase inhibition.60 This would mean that growth factor breakdown would also be decreased, and an increase has been found.61 In later stages, the opioid as such inhibits maturation in animal models,62,63 and a complex nuclear magnetic resonance picture emerges, with decreases and increases in different structures.64 Of great interest are brain changes caused by inflammatory bowel disease (IBD) in areas most affected in autism.65

Panksepp66 showed that opioids induced social aloofness in animal models and especially abrogated social distress calls induced by removing the mother. The lack of socially meaningful relationships is also very pronounced in human opium/heroin addicts. Opioids and casomorphins have been shown to induce epileptic seizures, especially those of the temporal cortex.67 Epilepsy is common in autistic children,68 and contrary to controls, increases with age. Fluctuating analgesia has been registered in autism and would be expected in opioids that have a dietary origin.

That genetics and environment interact is not news, and if exposed to increased mercury (Hg) load that acts as an enzyme poison and measured as bioactive Hg,69 a dietary overload may well explain the increased frequency of autism in recent years. Also, in Norway, we have found an increase in functional Hg in autistic children (in preparation). In Norway, the increase in daily consumption of casein can be seen in TABLE 5. The considerable increase in milk-derived compounds has also been found in the United States, the land of ice cream, milk shakes, etc.


Change in milk products consumed in norway, causing increased dietary loading

  1946 1950 1960 1970 1980 1985 1986 % change
Milk, million L 369 393 488 811 880 799 738 100
Butter, tons 6393 11,206 18,069 17,584 21,077 24,286 23,182 262.6
White cheese, tons 4634 15,873 25,896 36,983 52,566 52,670 57,169 1133.7
Brown cheese, tons 5021 13,416 13,844 15,296 16,914 16,573 16,017 219
Melted cheese, tons 2836 2258 2948 1.06
Conserved milk products, L 25,747 24,669 27,472 27.9
Population, millions 3.157 3.280 3.571 3.866 4.979 4.146 4.159 31.7
The data were supplied by Tine Milk Dairy Co. The data show that the increase in milk-derived products has changed considerably, outstripping the growth in population. Percent change was calculated using 1946 as the starting point or when registered the first time (melted cheese, 1975, was 2917 tons, and conserved milk, 1975, was 21,485 L). In 1946, pizza was an unknown entity in most of Scandinavia.
Other evidence in support of a gut-brain connection

Interestingly, celiac disease seems to predispose to depression,70 and autistic traits are found in children with celiac disease.71 Chronic gastrointestinal problems are connected to a lifetime prevalence of depression.72-75 Finally, exposing children with ADHD and intolerance to certain foods76 showed magnetic EEG changes when exposed to the same foods.77

A series of articles have documented CNS damage with increased gliadin uptake, including peripheral neuropathy and cerebellar ataxia,78 and epilepsy.79 These findings have been extensively discussed.34


The evidence for a gut-to-brain connection is considerable. Physiology and the experimental facts show that this is not only plausible but probable.34,80 Medical diagnoses, which are largely based on symptoms only, usually are quite heterogenous. Disorders like fragile X, tuberous sclerosis, and Klüver-Bucy syndrome do not show peptide increases, and there are some unexplained cases as well. Measurable behavioral changes have been obtained in children with autism when gluten and casein are removed.

DISCLOSURES: Dr. Reichelt has received grant/research support from the Autism Research Institute, San Diego, CA, and is an unpaid consultant to Biomedical Laboratory in Oslo, Norway. Dr. Knivsberg reports no financial relationship with any company whose products are menioned in this article or with manufacturers of competing products.

ACKNOWLEDGEMENTS: We thank Dr. B. Rimland for supporting our work financially and with his great inspiration and friendship.


  1. Folstein H, Rutter M. Infantile autism: a genetic study of 21 twin pairs. J Child Psychol Psychiatry. 1977;18:297–321.
  2. Bailey A, Le Couteur A, Gottesman I, et al. Autism as a strongly genetic disorder: evidence from a British twin study. Psychol Med. 1995;25:63–77.
  3. Ames BN, Elson-Schwab I, Silver EA. High-dose vitamin therapy stimulates variant enzymes with decreased coenzyme binding affinity (increased K(m)): relevance to genetic disease and polymorphism. Am J Clin Nutr. 2002;75:616–658.
  4. Svedberg J, de Haas J, Leimenstoll G, et al. Demonstration of beta-casomorphin immunoreactive materials in vitro digests of bovine milk and in small intestinal contents after bovine milk ingestion in adult humans. Peptides. 1985;6:825–830.
  5. Chabance B, Marteau P, Rambaud JC, et al. Casein peptide release and passage to the blood in humans during digestion of milk and yogurt. Biochimie. 1998;80:155–165.
  6. Gardner MLG. Absorption of intact proteins and peptides. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York, NY: Raven Press;1994:1795–1820.
  7. Mahe S, Tome D, Dumontier AM, et al. Absorption of intact morphiceptin by diisopropylflurophosphate-treated rabbit ileum. Peptides. 1989;10:45–52.
  8. Bouras M, Huneau JF, Tomé D. The inhibition of intestinal dipeptidylaminopeptidase-IV promotes the absorption of enterostatin and des-arginine-enterostatin across rat jejunum in vitro. Life Sci. 1996;59:2147–2155.
  9. Husby S, Jensenius JC, Cant AJ. Passage of undegraded dietary antigen into the blood of healthy adults. Scand J Immunol. 1985;22:83–92.
  10. Seifert J, Ganser R, Brebdel W. Absorption of a proteolytic enzyme originating from plants out of the gastrointestinal tract into blood and lymph of rats [in German]. Z Gastroenterol. 1979;17:1–8.
  11. Scott H, Rognum TO, Midtvedt TJ, et al. Age-related changes in human serum antibodies to dietary and colonic bacteria antigens measured by enzyme-linked immunoabsorbeant assay. Acta Pathol Microbiol Immunol Scand C. 1985;93:65–70.
  12. Kilshaw PJ, Cant AJ. The passage of maternal dietary protein into human breast milk. Int Arch Allergy Appl Immunol. 1984;75:8–15.
  13. Stuart CA, Twiselton R, Nicholas MK, et al. Passage of cow’s milk protein in breast milk. Clin Allergy. 1984;14:533–535.
  14. Axelsson I, Jacobsson I, Lindberg T, et al. Bovine lactoglobulin in human milk. A longitudinal study during the whole lactation period. Acta Paediatr Scand. 1986;75:702–707.
  15. Troncone R, Scarcello A, Donatiello A, et al. Passage of gliadin into human breast milk. Acta Paediatr Scand. 1987;76:453–456.
  16. Langer SZ. 25 years since the discovery of presynaptic receptors: present knowledge and future perspectives. Trends Pharmacol Sci. 1997;18:95–99.
  17. Reichelt WH, Ek J, Stensrud MB, et al. Peptide excretion in celiac disease. J Pediatr Gastroenterol Nutr. 1998;26:305–309.
  18. Jyonouchi H, Sun S, Itokazu N. Innate immunity associated with inflammatory responses and cytokine production against common dietary proteins in patients with autism spectrum disorder. Neuropsychobiology. 2002;46:76–84.
  19. Torrente F, Ashwood P, Day R, et al. Small intestinal enteropathy with epithelial IgG and complement deposition in children with regressive autism. Mol Psychiatry. 2002;7:375–382.
  20. Ashwood P, Anthony A, Pellicer AA, et al. Intestinal lymphocyte populations in children with regressive autism: evidence for extensive mucosal immunopathology. J Clin Immunol. 2003;23:504–517.
  21. Horvath K, Papadimitriou JC, Rabasztyn A, et al. Gastrointestinal abnormalities in children with autistic disorder. J Pediatr. 1999;135:559–563.
  22. Heyman M, Desjeux JF. Cytokine-induced alteration of epithelial barrier to food antigens in disease. Ann N Y Acad Sci. 2000;915:304–311.
  23. Ermisch A, Rühle HJ, Neubert K, et al. On the blood-brain barrier to peptides: [3H]beta-casomorphin-5 uptake by eighteen brain regions in vivo. J Neurochem. 1983;41:1229–1233.
  24. Nyberg F, Lieberman H, Lindström LH, et al. Immunoreactive beta-casomorphin-8 in cerebrospinal fluid from pregnant and lactating women: correlation with plasma level. J Clin Endocrinol Metab. 1989;68:283–289.
  25. Sun Z, Cade JR. A peptide found in schizophrenia and autism causes behavioral changes in rats. Autism. 1999;3:85–95.
  26. Sun Z, Cade RJ, Fregly MJ, et al. Beta-casomorphin induces Fos-like immunoreactivity in discrete brain regions relevant to schizophrenia and autism. Autism. 1999;3:67–83.
  27. Fanciulli G, Dettori A, Fenude E, et al. Intravenous administration of food-derived opioid peptide Gluten Exorphin B5 stimulates prolactin secretion in rats. Pharmacol Res. 2003;47:53–58.
  28. Hole K, Bergslien H, Jørgensen HA, et al. A peptide containing fraction from schizophrenia which stimulates opiate receptors and inhibits dopamine uptake. Neuroscience. 1979;4:1883–1893.
  29. Drysdale A, Deacon R, Lewis P, et al. A peptide-containing fraction of plasma of schizophrenic patients which binds to opiate receptors and induces hyperactivity in rats. Neuroscience. 1982;7:1567–1573.
  30. Reichelt KL, Ekrem J, Scott H. Gluten, milk proteins and autism: dietary intervention effects on behavior and peptitde secretion. J Appl Nutr. 1990;42:1–11.
  31. Shattock P, Kennedy A, Rowell F, et al. Role of neuropeptides in autism and their relationship with classical neuro-transmitters. Brain Dysfunction. 1990;3:328–345.
  32. Cade R, Privette M, Fregly M, et al. Autism and schizophrenia: intestinal disorders. Nutr Neurosci. 2000;3:57–72.
  33. Shanahan MR, Venturini AJ, Daiss JL, et al. Peptide diagnostic markers for human disorders. European patent application. 2000;EP 0 969 015 A2:1–44.
  34. Reichelt KL, Knivsberg AM. Can the pathophysiology of autism be explained by the nature and the discovered urine peptides? Nutr Neurosci. 2003;6:19–28.
  35. Kawashti MI, Amin OR, Rowehy NG. Possible immunological disorders in autism: concomitant autoimmunity and immune tolerance. Egypt J Immunol. 2006;13:99–104.
  36. Wakefield AJ, Anthony A, Murch SH, et al. Enterocolitis in children with developmental disorders. Am J Gastroenterol. 2000;95:2285–2295.
  37. Horvath K, Perman JA. Autism and gastrointestinal symptoms. Curr Gastroenterol Rep. 2002;4:251–258.
  38. D’Eufemia P, Celli M, Finnochiaro R, et al. Abnormal intestinal permeability in children with autism. Acta Paediatr. 1996;85:1076–1079.
  39. Paul KD, Henker J, Todt A, et al. EEG-befunde zoeliakikranken kindern in abhängigkeit von der ernährung. Zeitschrift fur Klinische Medizin. 1985;40:707–709.
  40. Geissler A, Andus T, Roth M, et al. Focal white-matter lesions in brain of patients with inflammatory bowel disease. Lancet. 1995;345:897–898.
  41. Hart PE, Gould SR, MacSweeney JE, et al. Brain white-matter lesions in inflammatory bowel disease. Lancet. 1998;351:1558–1558.
  42. Blau N, Niederwieser A, Shmerling DH. Peptiduria presumably caused by aminopeptidase-P deficiency: a new error of metabolism. J Inherit Metab Dis. 1980;11(suppl 2):240–242.
  43. Abassi Z, Golomb E, Keiser HR. Neutral endopeptidase inhibition increases urinary excretion and plasma level of endothelin. Metabolism. 1992;41:683–685.
  44. Watanabe Y, Kojima-Komatsu T, Iwaki-Egawa S, et al. Increased excretion of proline-containing peptides in dipeptidyl peptidase IV deficient rats. Res Commun Chem Pathol Pharmacol. 1993;81:323–330.
  45. Undrum T, Lunde HA, Gjessing LR. Determination of ophidine in human urine. J Chromatogr. 1982;227:53–59.
  46. Sponheim E, Myhre AM, Reichelt KL, et al. Urine peptide patterns in children with milder types of autism [in Norwegian]. Tidsskr Nor Laegeforen. 2006;126:1475–1477.
  47. Knivsberg AM, Wiig K, Lind G, et al. Dietary intervention in autistic syndromes. Brain Dysfunction. 1990;3:315–327.
  48. Lucarelli S, Frediani T, Zingoni AM, et al. Food allergy and infantile autism. Panminerva Med. 1995;37:137–141.
  49. Knivsberg AM, Reichelt KL, Nødland M, et al. Autistic syndromes and diet: a follow-up study. Scandinavian Journal of Educational Research. 1995;39:223–236.
  50. Whiteley P, Rodgers J, Savery D, et al. A gluten-free diet as an intervention for autism and associated spectrum disorders: preliminary findings. Autism. 1999;3:45–65.
  51. Knivsberg AM, Reichelt KL, Høien T, et al. A randomized, controlled study of dietary intervention in autistic syndromes. Nutr Neurosci. 2002;5:251–261.
  52. Klaveness J, Bigam J. The GFCF Kids diet survey. In: The Autism Research Unit, Sunderland University. Sunderland, UK: Building Bridges. 2002;77–84.
  53. Rimland B. Parent rating of behavioral effects of biomedical interventions ARI publication 34. San Diego, CA: Autism Research Institute; 2003.
  54. Kniker WT, Andrews A, Hundley A, et al. The possible role of intolerance to milk/dairy and wheat/gluten foods in older children and adults with autism spectrum disorder. In: The Autism Research Unit, Sunderland University, Sunderland, UK. 2001: An autism odyssey. 2001:183–191.
  55. Brudnak MA, Rimland B, Kerry RE, et al. Enzyme-based therapy for autism spectrum disorders—is it worth another look? Med Hypotheses. 2002;58:422–428.
  56. Knivsberg AM, Reichelt KL, Nödland M. Dietary intervention for a seven year old girl with autistic behaviour. Nutr Neurosci. 1999;2:435–439.
  57. Lelord G, Muh JP, Berthelemey C, et al. Effects of pyridoxine and magnesium on autistic symptoms: initial observations. J Autism Dev Disord. 1981;11:219–230.
  58. Dolske MC, Spollen J, McKay S, et al. A preliminary trial of ascorbic acid as supplement therapy in autism. Prog Neuropsychopharmacol Biol Psychiatry. 1993;17:765–774.
  59. Courhesne E. Brain development in autism: early overgrowth followed by premature arrest of growth. Ment Retard Dev Disabil Res Rev. 2004;10:106–111.
  60. LaBella FS, Geiger JD, Glavin GB. Administered peptides inhibit the degradation of endogenous peptides. The dilemma of distinguishing direct from indirect effects. Peptides. 1985;6:645–660.
  61. Nelson KB, Grether JK, Croen LA, et al. Neuropeptides and neurotrophins in neonatal blood of children with autism or mental retardation. Ann Neurol. 2001;49:597–606.
  62. Hauser KF, McLaughlin PJ, Zagon IS. Endogenous opioid systems and the regulation of dendritic growth and spine formation. J Comp Neurol. 1989;281:13–22.
  63. Zagon IS, McLaughlin PJ. Endogenous opioid systems regulate cell proliferation in the developing rat brain. Brain Res. 1987;412:68–72.
  64. Hashimoto T, Tayama M, Miyazaki M, et al. Reduced brain stem size in children with autism. Brain Dev. 1992;14:94–97.
  65. Welch MG, Welch-Horan TB, Anwar M, et al. Brain effects of chronic IBD in areas abnormal in autism and treatment by single neuropeptides secretin and oxytocin. J Mol Neurosci. 2005;25:259–274.
  66. Panksepp J, Normansell L, Sivily S, et al. Casomorphins reduce separation distress in chicks. Peptides. 1984;5:829–831.
  67. Siggins GR, Henriksen SJ, Chavkin CA, et al. Opioid peptides and epileptogenesis in the limbic system: cellular mechanisms. Adv Neurol. 1986;44:501–512.
  68. Deykin EY, MacMahon B. The incidence of seizures among children with autistic symptoms. Am J Psychiatry. 1995;136:1310–1312.
  69. Nataf R, Skorupka C, Amet L, et al. Porphyrinuria in childhood autistic disorder: implications for environmental toxicity. Toxicol Appl Pharmacol. 2006;214:99–108.
  70. Hallert C, Åstrøm J, Sedvall G. Psychic disturbances in adult coeliac disease. III. Reduced monoamine metabolism and signs of depression. Scand J Gastroenterol. 1982;17:25–28.
  71. Asperger H. Psychopathology of children with coeliac disease [in German]. Ann Paediatr. 1961;197:346–351.
  72. Lydiard RB, Fossey MD, Marsh W, et al. Prevalence of psychiatric disorders in patients with irritable bowel syndrome. Psychosomatics. 1993;34:229–234.
  73. Addolorato G, Capristo E, Stefanini GF, et al. Inflammatory bowel disease: a study of the association between anxiety and depression, physical morbidity, and nutritional status. Scand J Gastroenterol. 1997;32:1013–1021.
  74. Haug TT, Mykletun A, Dahl AA. Are anxiety and depression related to gastrointestinal symptoms in the general population? Scand J Gastroenterol. 2005;37:294–298.
  75. Alander T, Svärdsudd K, Johansson SE, et al. Psychological illness is commonly associated with functional gastrointestinal disorders and is important to consider during patients consultations: a population-based study. BMC Med. 2005;3:8–8.
  76. Egger J, Carter CM, Graham PJ, et al. Controlled trial of oligoantigenic treatment in the hyperkinetic syndrome. Lancet. 1985;1:540–545.
  77. Uhlig T, Merkenschlager A, Brandmaier R, et al. Topographic mapping of brain electrical activity in children with food-induced attention deficit hyperactivity disorder. Eur J Pediatr. 1997;156:557–561.
  78. Hadjivassiliou M, Boscolo S, Davies-Jones GA, et al. The humoral response in the pathogenesis of gluten ataxia. Neurology. 2002;58:1221–1226.
  79. Gobbi G, Bouquet F, Greco L, et al. Coeliac disease, epilepsy, and cerebral calcifications. Lancet. 1992;340:439–443.
  80. Wakefield AJ, Puleston JM, Montgomery SM, et al. Review article: the concept of entero-colonic encephalopathy, autism and opioid receptor ligands. Aliment Pharmacol Ther. 2002;16:663–674.

CORRESPONDENCE: K. L. Reichelt, MD, PhD, Department of Pediatric Research, Rikshospitalet, N0027, Oslo, Norway E-MAIL: