Celiac disease (CD) is an inflammatory disease of the upper small intestine resulting from gluten ingestion in genetically susceptible individuals. The immunologically based inflammation causes atrophy of the villous structure of the jejunum, leading to malabsorption of several important nutrients.¹ Hereditary factors alone do not explain the development of CD, although 90% of patients share the HLA-DR3/DQ2 haplotype. Immunological factors, infectious agents, and hormonal status may also be involved.¹ The terms "gluten sensitivity" and "latent CD" refer to states of heightened immunological responsiveness to ingested gluten in individuals with susceptibility genes for CD, but without small-bowel villous atrophy.^1,2

During the last 20 years it has become evident that CD is underdiagnosed. In a recent screening study of a healthy population, the prevalence of CD was as high as 1 in 130.³ Clinical CD may present at any age. Minor symptoms such as fatigue, dyspepsia, anemia, or slight weight loss have been found to be increasingly common. The disease may even be clinically silent, despite manifest jejunal lesion. Diagnosis is based on demonstration of villous atrophy in a biopsy of the jejunum, and a gluten-free diet is the treatment of choice.¹

Neurological symptoms such as ataxia are common in CD and may be the sole manifestation of disease in some individuals with genetic susceptibility; even with serological evidence of gluten sensitivity, there may be no gastrointestinal symptoms or manifest mucosal lesion.² Besides neurological symptoms, psychiatric disorders have been found to be prevalent in untreated CD. One study⁴ reported a prior history of psychiatric treatment in a high proportion of adults with CD, even years before the CD diagnosis; 21% (9 of 42) had attended a psychiatric clinic because of "neurotic, mostly depressive disorders," and six had been granted a disability pension for chronic depression. Only 5% (2 of 42) of the medical comparison group patients in this study had a history of psychiatric treatment. Using a modified version of the Zung-Self-Rating Depression Scale, Ciacci et al.⁵ found that significantly more CD patients (about one-third of 92 CD patients) had depressive symptoms, compared with patients with chronic persistent hepatitis and with normal subjects. These findings suggest an increased prevalence of mental symptoms among adults with CD. However, as none of these studies applied current diagnostic criteria or structured psychiatric interviews, the validity of the diagnoses remains unclear.

The mechanisms involved in the etiology and pathogenesis of mental and behavioral disorders related to CD are unclear. Hallert and Sedvall⁶ reported an increase of 33% in major monoamine metabolite (5-hydroxyindoleacetic acid [5-HIAA], homovanillic acid, and 3-methoxy-4-hydroxyphenylglycol) concentrations and a 10% increase in the concentration of tryptophan in the cerebrospinal fluid (CSF) in adult CD patients after 1 year on a gluten-free diet. Hernanz and Polanco⁷ found significantly decreased plasma concentrations of tryptophan, citrulline, tyrosine, valine, isoleucine, and leucine and significantly diminished ratios of tryptophan to large neutral amino acids in children with CD, regardless of dietary treatment. In untreated children, plasma tryptophan was 84% less (mean±SD=13±4 µmol/L) and in the treated group 62% less (31±3 µmol/L) than in children without CD (81±22 µmol/L), and the plasma tryptophan ratio was found to be significantly lower in the untreated group compared with the treated and control groups. Nine of 15 children with untreated CD showed signs of "behavioral disturbances" and were irritable or apathetic. In some of these patients, mood and behavioral problems improved after starting a gluten-free diet. However, to our knowledge, these findings have not been confirmed elsewhere.

To our knowledge, there are no studies based on structured psychiatric interviews that have examined mental disorders associated with CD in children and adolescents. We describe two adolescent patients who suffered from severe mental and behavioral disorders before receiving a CD diagnosis. In both cases, psychiatric status considerably improved soon after the commencement of a gluten-free diet.

Case Report

Case 1. Anne was the first of three children in her family. Rheumatic disease and dysthymic disorder had been diagnosed in a family member on the mother's side but not among Anne's first-degree relatives. At the age of 1.5 years she was diagnosed with lactose intolerance, and at age 7 investigations for mild arthritis revealed anemia and an elevated level of antigliadin antibodies. A subsequent jejunal biopsy was normal. After she began school she had occasional abdominal pains. From age 11 years she avoided fat in foods and ate a very restricted diet. At age 13.5 years Anne had not yet begun menstruating, although her mother had begun at age 11. Anne was 164 cm tall, weighed 46.2 kg (12% less than that expected), and desperately wanted to weigh less to be "perfect." Anorexia nervosa was suspected. Investigations for her underweight and recurrent abdominal pains revealed anemia (hemoglobin of 114 mg/ L) and elevated levels of antigliadin, IgA antiendomysium, and IgA antitransglutaminase antibodies. Jejunal biopsy revealed villous atrophy, confirming the diagnosis of CD when Anne was 13.5 years of age. A gluten-free diet was instituted. In 6 months the level of IgA antitransglutaminase antibodies had normalized (from 52% to 5%), and the level of IgA antiendomysium antibodies had decreased from 400 to 100 U/mL, indicating a positive response to the gluten-free diet.

Since the age of 12, contrary to her previous behavior, Anne had serious problems controlling her temper and problems with explosive anger. From age 13 years (6 months before the CD diagnosis), she felt continuously tired, depressed, and irritated. She lost interest and pleasure in activities. She slept more but was tired and lacked energy and had difficulties thinking, concentrating, and making decisions. Her school performance deteriorated. Anne felt ugly, disgusting, and worthless and was prone to feelings of rejection and inappropriate guilt. She had always been shy, reserved, and lonely, but she began to withdraw even further from her family.

During the final 6 months before the diagnosis of CD, Anne met the criteria for major depressive disorder, moderate, single episode, and eating disorder not otherwise specified (meeting all the criteria for anorexia nervosa, except for weight, as her relative weight was 12% less than expected), on the basis of an interview with the Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL).^8,9 She scored 23 on the Beck Depression Inventory (BDI),¹⁰ indicating moderate depression. Her scores on the Child Behavior Checklist (CBCL)¹¹ indicated that her withdrawal, thinking problems, and aggressive behavior were clinically abnormal.

Within 2 months after Anne started a gluten-free diet, her depressive symptoms diminished, without psychiatric treatment or known psychosocial factors to explain her remission. She scored 0 on the BDI. Her abdominal pains and tiredness subsided, and her weight began to increase slowly (her relative weight was then 10% less than expected) without causing excessive anxiety. Fifteen months later she was worried about her mood swings and loneliness, but she did not meet the criteria for any current mental disorder. Her scores on the CBCL indicated that only the social problems remained at the borderline level. She was 170.3 cm tall and weighed 54 kg (4% less than expected). She had begun menstruating, and her adolescent development forged ahead. She remained in remission throughout the follow-up period of 2 years.

Case 2. Tom was the first and only child in his family. At the age of 3 years he was suspected of having an autistic-like contact disorder because he was uncommunicative in kindergarten and talked to himself using a self-made language. Although his first grade schoolteacher insisted that Tom see a child psychiatrist because of his disruptive restlessness, no consultation took place. Tom recalled that his school years were largely dismal and depressive. He had difficulties concentrating and studying and was shunned by his schoolmates.

Tom was growing normally and never suffered abdominal complaints. At age 14.5 years, when he was 174.2 cm tall (1.2 SD taller than expected) and weighed 64.7 kg (9% more than expected), he underwent laboratory tests because both of his parents had received the diagnosis of CD 5 years earlier. The results revealed slightly elevated antigliadin antibodies and markedly elevated IgA antiendomysium antibodies (>400 U/mL), but the results of a jejunal biopsy did not meet the criteria for CD. One year later, when Tom was 15.5 year old, his mother insisted on a second jejunal biopsy because of his severe and increasing tiredness. The results showed subtotal villous atrophy with crypt hyperplasia, confirming the diagnosis of CD. A gluten-free diet was started, and in 6 months the levels of IgA antiendomysium and IgA antitransglutaminase antibodies normalized, indicating a positive response to the gluten-free diet. One year later a low titer of IgA antiendomysium antibodies was found (titer 1:50), and the level of serum folate remained subnormal (5.3 nmol/ L).

A psychiatric examination of Tom was made when he was age 17, 1.5 years after the diagnosis of CD. When Tom's mother was interviewed, she said that she had been depressed after Tom's birth and also that Tom's father suffered from depressive disorder and shortness of temper, which were also common symptoms in his relatives. By means of a psychiatric research interview based on the K-SADS-PL, Tom was retrospectively found to have met the diagnostic criteria for dysthymic disorder between age 7 and 16 years. After age 8, he had more prominent recurrent depressive episodes, indicating double depression. From age 11 to 13 years he often lost his temper at school and occasionally hurt others, consequently satisfying the criteria for intermittent explosive disorder. The criteria for obsessive-compulsive disorder with obsessive aggression and sexually-related thoughts were met from age 13 years onward. During the 6 months before the diagnosis of CD at age 15.5 years, he met criteria for major depressive episode, severe, with transient psychotic symptoms. Tom could not sleep until dawn and was restless and terrified by his own thoughts. He was tired, often late for school, and frequently stayed alone at home. He could not concentrate on studying and suffered from anhedonia, indecisiveness, and inappropriate feelings of guilt and worthlessness. Tom slashed his arm superficially several times but denied suicidal intent. Because Tom did not communicate his psychiatric symptoms to anyone, his mental disorders remained unrecognized and he did not have specific psychiatric treatment.

In the 5 months after starting a gluten-free diet, Tom gradually remitted from the major depressive episode without psychiatric treatment or known psychosocial factors to explain the remission. Within 2 months his severe sleeping problems subsided, and he was able to attend school normally. His school performance improved. He learned how to control his obsessive thoughts and fears, approaching them "on a philosophical level," and he felt able to believe in himself after 9 years of depression. At age 17 Tom did not meet the diagnostic criteria for any current mental disorder, although he had serious difficulties in establishing contact with others. He scored 6 on the BDI, and on the CBCL completed by the parent only his T-score for withdrawal was clinically abnormal. Tom felt that his self-esteem was beginning to improve and that he was coping better with his chronic loneliness.

Discussion

The two adolescents described in these cases had suffered from episodes of major depression and other mental disorders before receiving a diagnosis of CD. The subject in case 2 had severe psychiatric symptoms years before adolescence. Soon after commencement of a gluten-free diet, coinciding with a decrease in circulating IgA antiendomysium and IgA antitransglutaminase antibodies, both youngsters considerably improved without any specific psychiatric treatment, and both remained in remission for at least 1.5 years of follow-up. Although the possible role of unrecognized psychosocial factors in explaining the remission cannot be excluded, it seems likely that in these cases major depression and severe behavioral problems, along with their improvement, were causally related to CD and its treatment with a gluten-free diet.

There are several possible pathways through which CD can influence the central nervous system and predispose the CD patient to mental disorders. Malabsorption of important nutrients such as folic acid, vitamin B₆ and amino acids, especially tryptophan, may lead to disturbances in CNS serotonin function associated with major depressive disorder and aggressive behavior. The mood-lowering effect of experimental rapid tryptophan depletion has been demonstrated, for example, in medication-free women with recurrent depressive episodes.¹² Decreased tryptophan levels found in untreated children with CD⁷ are similar to those found in tryptophan depletion experiments. According to van Praag¹³ the serotonergic disturbances found in some depressed individuals, particularly those with lower levels of CSF 5-HIAA, are linked to the anxiety and aggression components of depressive symptoms and to heightened sensitivity to stressful life events. Serotonin function is not linearly related to the level of depression, and the exact nature of serotonergic impairment is still unknown. Delgado et al.¹⁴ have suggested that lower levels of serotonin function could feature in the pathophysiology of depression, either as a predisposing factor or as an effect of a postsynaptic deficit in serotonin utilization.

The role of immune activation in the pathogenesis of CD is widely accepted.¹ "Gluten sensitivity" and "latent CD" refer to states of heightened immunological responsiveness to ingested gluten in genetically susceptible individuals with normal small bowel mucosal morphology.^1,3 Lahat et al.¹⁵ studied the cytokine profile in active CD and confirmed enhanced production of interferon-gamma (IFN-). Maes and Smith¹⁶ have proposed excessive cytokine secretion due to chronic immune system activation as a fundamental pathology underlying depressive symptoms. Cytokines as such cannot cross the blood-brain barrier, but growing evidence suggests that specific cytokines may signal the brain to produce neurochemical, neuroendocrine, neuroimmune, and behavioral changes.¹⁷ Cytokine activation is known to enhance the hypothalamus-pituitary-adrenal axis hyperactivity associated with major depression.¹⁸ Increased IFN- production can suppress serotonin function both directly and indirectly, such as by enhancing tryptophan as well as serotonin turnover by means of increased activity of the kynurenine-niacin pathway (see, for example, Brown¹⁹). Up to 90% suppression of plasma tryptophan levels may be produced by interferon administration²⁰ or by experimental tryptophan depletion.²¹ Any such mechanism could be operative in untreated CD and could cause disturbances in brain serotonin function, predisposing the patient to mood and behavioral disorders.

Since unrecognized CD may predispose the sufferer to serious mental disorders and behavioral problems, it should be taken into account in differential diagnosis in all age groups. The mechanisms involved in the etiology and pathogenesis of mental and behavioral disorders related to CD, and even to celiac-type gluten sensitivity, remain unresolved.

ACKNOWLEDGMENTS

Preparation of this report was supported by a grant from the Yrjö Jahnsson Foundation.

REFERENCES

1. Mäki M, Collin P: Coeliac disease. Lancet 1997; 349:1755-1759

2. Hadjivassiliou M, Grunewald RA, Chattopadhyay AK, et al: Clinical, radiological, and neuropathological characteristics of gluten ataxia. Lancet 1998; 352:1582-1585

3. Kolho KL, Färkkilä MA, Savilahti E: Undiagnosed coeliac disease is common in Finnish adults. Scand J Gastroenterol 1998; 33:1280-1283

4. Hallert C, Derefeldt T: Psychic disturbances in adult coeliac disease: 1. clinical observations. Scand J Gastroenterol 1982; 17:17-19

5. Ciacci C, Iavarone A, Mazzacca G, et al: Depressive symptoms in adult coeliac disease. Scand J Gastroenterol 1998; 33:247-250

6. Hallert C, Sedvall G: Improvement in central monoamine metabolism in adult coeliac patients starting a gluten-free diet. Psychol Med 1983; 13:267-271

7. Hernanz A, Polanco I: Plasma precursor amino acids of central nervous system monoamines in children with coeliac disease. Gut 1991; 32:1478-1481
8. Aronen ET, Mäkelä J, Cacciatore R, et al: K-SADS-PL, Finnish Translation. Helskinki, Hospital for Children and Adolescents, University of Helsinki, 1998

9. Kaufman J, Birmaher B, Brent D, et al: Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL): initial reliability and validity data. J Am Acad Child Adolesc Psychiatry 1997; 36:980-988

10. Beck AT, Ward CH, Mendelson M, et al: An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561-571

11. Achenbach TM: Manual for the Child Behavior Checklist 4-18 and 1991 Profile. Burlington, University of Vermont, Department of Psychiatry, 1991

12. Smith KA, Fairburn CG, Cowen PJ: Relapse of depression after rapid depletion of tryptophan. Lancet 1997; 349:915-919

13. Van Praag HM: Anxiety and increased aggression as pacemakers of depression. Acta Psychiatr Scand Suppl 1998; 393:81-88

14. Delgado P, Price L, Miller H, et al: Serotonin and the neurobiology of depression: effects of tryptophan depletion in drug-free depressed patients. Arch Gen Psychiatry 1994: 51:865-874

15. Lahat N, Shapiro S, Karban A, et al: Cytokine profile in coeliac disease. Scand J Immunol 1999; 49:441-446

16. Maes M, Smith R: Immune activation and major depression: a hypothesis. Perspectives in Depression 1999; 7:6-8

17. Kronfol Z, Remick DG: Cytokines and the brain: implications for clinical psychiatry. Am J Psychiatry 2000; 157:683-694

18. Maes M, Smith R, Scharpe S: The monocyte-T-lymphocyte hypothesis of major depression. Psychoneuroendocrinology 1995; 20:111-116

19. Brown RR: Metabolism and biology of tryptophan, in Recent Advances in Tryptophan Research: Tryptophan and Serotonin Pathways. Edited by Filippini GA, Costa CVL, Bertazzo A. New York, Plenum, 1996, pp 15-25

20. Strander H: Toxicities of interferons, in Clinical Applications of the Interferons, edited by Stuart-Harris R, Penny R. London, Chapman & Hall, 1997, pp. 331-363

21. Menkes D, MacDonald J: Interferons, serotonin and neurotoxicity. Psychol Med 2000; 30:259-268