Effect of a gluten free diet on osteopenia in adults with newly diagnosed coeliac disease.

BACKGROUND/AIMS: Calcium and vitamin D malabsorption in coeliac disease predispose to skeletal demineralisation. This study aimed to determine bone mineral density in patients studied in the first year after diagnosis of coeliac disease, and to detect changes in bone mineral density over the subsequent year. METHODS: Lumbar spine and femoral neck bone mineral density was measured in 21 adults with coeliac disease, diagnosed and started on a gluten free diet during the preceding year, with dual energy x ray absorptiometry and repeated after 12 months. RESULTS: Bone mineral density was significantly lower in patients than in paired controls (matched for age and sex), at lumbar spine (0.819 g/cm2 compared with 1.021 g/cm2, p < 0.001 Wilcoxon signed rank test) and femoral neck (0.663 g/cm2 compared with 0.794 g/cm2, p < 0.001). Repeat measurement after 12 months demonstrated that patients had a significant gain in bone mineral density at lumbar spine (16.6%/year), and femoral neck (15.5%/year, p < 0.002, Wilcoxon signed rank test at both sites), whereas no significant change in bone mineral density was detected in controls. CONCLUSIONS: Treatment of coeliac disease with a gluten free diet is associated with a significant increase in bone mineral density, although patients still had lower bone mineral density than controls.     

Detection of low bone mineral density by dual energy x ray absorptiometry in unsuspected suboptimally treated coeliac disease.

Patients with coeliac disease may present with calcium malabsorption but it is unclear whether this results in longterm impairment of bone mineralisation. Dual energy x ray absorptiometry (DXA) was used to study bone mineral density in 34 asymptomatic coeliac disease patients, treated with a gluten free diet for at least two years, and also in 10 newly diagnosed or untreated patients. As expected, untreated patients had low bone mineral density in all regions. In the 29 treated female coeliac disease patients, overall mean values for age adjusted bone mineral density expressed as Z scores were normal although there were many patients with low values, particularly of the lumbar spine and total body. Scores in the postmenopausal patients were significantly worse than in the premenopausal patients and low mean Z scores were found in the five treated male patients. The subjects who had reduced bone mineral density could not be predicted clinically but, despite being asymptomatic, were more likely to have subtotal or partial villous atrophy on small intestinal biopsy (p < 0.0275). In conclusion, although many treated coeliac disease patients have normal bone mineral density, suboptimally treated and newly diagnosed or untreated patients have osteopenia. To reduce the risk of osteoporotic fractures, it is recommended that bone mineral density be measured in all treated coeliac disease patients and those with osteopenia have a repeat intestinal biopsy to assess disease activity.     

Coeliac disease and bone mineral density in adult female patients.

A cross sectional study was undertaken to examine the relationship between coeliac disease and bone mineral density. The 135 female coeliac patients registered on the database of the Department of Gastroenterology at Hull Royal Infirmary were approached by letter, advising them of a potential risk of osteoporosis and inviting them to undergo bone densitometry. A total of 81 registered women (60%) attended the Osteoporosis Laboratory, Princess Royal Hospital and underwent dual energy x ray absorptiometry at the lumbar spine (L2-L4) and femoral neck. Historical data relating to the time of diagnosis and adherence to a gluten free diet were obtained. A control group was selected from the local normal population and was first matched for height, weight, and menopausal status. Postmenopausal patients were then further matched to controls of equivalent menopausal age. In coeliac patients, bone mineral density expressed in g/cm2 as mean (SD) was significantly lower at the lumbar spine (1.076 (0.186)) than in the control group (1.155 (0.143), p < 0.001). This was also the case at the femoral neck (0.887 (0.142) versus 0.965 (0.127), p < 0.001). When the coeliac patients were stratified by menopausal status, it was found that femoral neck bone mineral density was significantly below control values in both premenopausal and postmenopausal women. Spinal bone mineral density exhibited a significant decrement only in the postmenopausal group. The age at diagnosis of coeliac disease and adherence to a gluten free diet did not influence bone mineral density at either hip or spine. These results confirm coeliac patients' higher risk of osteopenia. Coeliac disease should be added to the list of medical conditions which constitute an indication for bone densitometry in order that the individual risk of osteoporosis related fracture may be determined.     

Osteoporosis in treated adult coeliac disease.

Forty five women and 10 men with coeliac disease diagnosed in adult life, who were already on a gluten free diet, had serial bone mineral density measurements at the lumbar spine and femoral neck over 12 months. Osteoporosis, defined as a bone mineral density (BMD) < or = 2 SD below the normal peak bone mass was found in 50% of male and 47% of female coeliac patients. Patients with a BMD < or = 2 SD below age and sex matched normal subjects, had a significantly lower body mass index (21.3 kg.m-2 compared with 25.2 kg.m-2, p < 0.02 Wilcoxon rank sum test) and lower average daily calcium intake (860 mg/day compared with 1054 mg/day, p < 0.05 Wilcoxon rank sum test) than patients with normal bone mineral density. In postmenopausal women with coeliac disease there was a strong correlation between the age at menopause and BMD at both the lumbar spine (r = 0.681, p < 0.01, Spearman's rank correlation) and femoral neck (r = 0.632, p < 0.01). No overall loss of bone was shown over the 12 months of follow up, and relative to the reference population there was a significant improvement in BMD at the lumbar spine in women (p < 0.025, paired t test) and at the femoral neck in men (p < 0.05, paired t test). There was a significant negative correlation between the annual percentage change in BMD at the lumbar spine and the duration of gluten free diet (r = -0.429, p<0.01, Spearman's rank correlation), with the largest gain in BMD in patients with most recently diagnosed coeliac disease. Osteoporosis was shown in 47% of patients with treated adult coeliac disease. Recognised risk factors for osteoporosis in the general population including low body mass index, dietary calcium intake, and early menopause are particularly important in coeliac disease. Treatment of coeliac disease with a gluten free diet probably protects against further bone loss, and in the early stages is associated with a gain in bone mineral density.     

Splenic atrophy in adult coeliac disease: is it reversible?

A study of splenic function in 28 patients with adult coeliac disease showed no significant correlation between the half life of heat-damaged red cells and either the duration of pre-treatment exposure to gluten or the length of time on a gluten free diet. A significant correlation was found between splenic size and duration of treatment; those patients who had been taking a gluten free diet for the longest time had the smallest spleens. Blood films from 11 of these 28 patients taken before treatment with a gluten free diet were compared with those taken between two and 15 years after the start of treatment. There was no tendency for the hyposplenic changes to regress. In the majority, the changes became more prominent despite strict adherence to the gluten free diet. These findings suggest that splenic atrophy in adult coeliac disease is not reversed by treatment with a gluten free diet and is unlikely to be related to the state of the jejunal mucosa or the duration of initial exposure to gluten.     

Cell-mediated immunity to gluten fraction III in adult coeliac disease

A migration inhibition test was used to assess sensitisation of blood leucocytes, and thus cell-mediated immunity, to gluten fraction III in controls and patients with coeliac disease. Migration indices were significantly less (indicating sensitisation) in untreated and in treated patients than in controls, and significantly less in treated patients than in untreated patients. At a concentration of 1 mg/ml gluten fraction III, 13% of untreated patients and 54% of treated patients had migration indices in the sensitised range. At 2 mg/ml gluten fraction III, sensitisation was demonstrated in 8% of untreated patients and 48% of treated patients. After starting a gluten free diet, migration indices fell into the sensitised range in all patients followed. After at least nine months on a gluten free diet, migration indices were significantly higher in those patients with a normal interepithelial lymphocyte count than in those patients with a raised interepithelial lymphocyte count. Cell-mediated immunity to gluten fraction III can be detected in the peripheral blood of certain patients with coeliac disease. Detectable sensitivity is related to the time on a gluten free diet, and the interepithelial lymphocyte count.     

Plasma enteroglucagon related to malabsorption in coeliac disease

Plasma enteroglucagon was measured before and during three hours after a standard meal in 21 untreated adult patients with suspected coeliac disease who all had villous atrophy of the small intestinal mucosa and malabsorption, and in nine control subjects. In 11 of these patients the diagnosis of coeliac disease was confirmed and 10 were again investigated on a gluten free diet. The coeliac patients had higher basal (37 +/- 9 pmol/l, mean +/- SE, p less than 0.05) and postprandial (70 +/- 9 pmol/l, p less than 0.005) mean plasma enteroglucagon concentrations than the control subjects (basal 14 +/- 4 pmol/l, postprandial 25 +/- 5 pmol/l). The 10 coeliac patients on gluten free diet for five to 20 months had a basal mean plasma enteroglucagon concentration not significantly lower than before treatment (25 +/- 5 pmol/l) but significantly lower postprandial enteroglucagon concentrations than before treatment (40 +/- 7 pmol/l, p less than 0.025). Postprandial plasma enteroglucagon concentration after 90 minutes in untreated patients correlated positively to the faecal fat excretion (r = 0.58, p less than 0.02). It correlated negatively to the urinary five hour D-xylose excretion after an oral load of 165 mmol D-xylose (r = -0.71, p less than 0.01). Thus, the postprandial plasma enteroglucagon concentrations in untreated coeliac disease were related to the degree of malabsorption and they normalised during treatment with a gluten free diet.     

Coeliac disease and unfavourable outcome of pregnancy

BACKGROUND: Up to 50% of women with untreated coeliac disease experience miscarriage or an unfavourable outcome of pregnancy. In most cases, after 6-12 months of a gluten free diet, no excess of unfavourable outcome of pregnancy is observed. The prevalence of undiagnosed coeliac disease among pregnant women is not known. AIM: To determine the prevalence of untreated coeliac disease among women attending the obstetrics-gynaecological department. METHODS: Endomysial antibodies, which are specific and sensitive for coeliac disease, were evaluated in all women attending the obstetrics-gynaecology department of a large city hospital over a 90 day period. RESULTS: Of 845 pregnant women screened, 12 were identified as having coeliac disease. Three had previously been diagnosed but were not following a gluten free diet. The remaining nine underwent a small intestinal biopsy, which confirmed the diagnosis. The outcome of pregnancy was unfavourable in seven of these 12 women. Six healthy babies were born with no problems after the women had been on a gluten free diet for one year. CONCLUSIONS: Overall, 1 in 70 women was affected by coeliac disease, either not diagnosed (nine cases) or not treated (three cases). Their history of miscarriages, anaemia, low birth weight babies, and unfavourable outcome of pregnancy suggests that testing for coeliac disease should be included in the battery of tests prescribed for pregnant women. Coeliac disease is considerably more common than most of the diseases for which pregnant women are routinely screened. Unfavourable events associated with coeliac disease may be prevented by a gluten free diet.     

Doppler ultrasonographic evaluation of splanchnic blood flow in coeliac disease.

BACKGROUND: Current knowledge on splanchnic haemodynamics in coeliac disease is limited and incomplete. AIM: To evaluate splanchnic arterial and venous blood flow in coeliac disease. METHODS: In 22 coeliac (13 untreated, nine treated) patients and in nine healthy subjects the following variables were assessed: vessel diameter and mean flow velocity in portal vein, splenic vein, superior mesenteric vein, and superior mesenteric artery. Peak systolic velocity, end diastolic velocity and pulsatility index were also determined in the superior mesenteric artery. Five patients of the untreated group were re-evaluated after nine months on a gluten free diet. RESULTS: Significant differences in haemodynamic variables between the three groups were shown only in the superior mesenteric artery. An increase in both mean flow velocity and end diastolic velocity and a reduction in pulsatility index occurred in untreated patients compared with treated patients (p < 0.002; p < 0.04; p < 0.035) and with healthy controls (p < 0.001; p < 0.025; p < 0.0003). Similar results were obtained for the five patients evaluated before and after treatment (p < 0.03; p < 0.02; p < 0.03), in whom the mean flow velocity in the superior mesenteric vein also decreased after treatment (p < 0.05). No significant differences were noted between treated coeliac patients and healthy controls. CONCLUSIONS: In untreated coeliac disease there is a hyperdynamic mesenteric circulation that decreases after treatment.     

Increase of lymphocytes bearing the gamma/delta T cell receptor in the jejunum of patients with dermatitis herpetiformis.

The densities of T cells and of cells bearing the T cell receptors gamma/delta and alpha/delta and the surface antigens CD4 and CD8 in jejunal specimens from 21 patients with dermatitis herpetiformis were compared with those in specimens from 13 untreated adults with coeliac disease and 13 control subjects. In the lamina propria of the jejunum the median density of gamma/delta+ cells was significantly (p less than 0.001) greater in untreated patients with dermatitis herpetiformis than in control subjects (114 v 36 cells/mm2) and similar to that found in the patients with coeliac disease (115 cells/mm2). The difference in gamma/delta+ cell density between patients with dermatitis herpetiformis and control subjects was much greater in the surface epithelium of the jejunum: the median density for 14 untreated patients with dermatitis herpetiformis was 39 cells/mm, for seven patients with dermatitis herpetiformis on a gluten free diet 34 cells/mm, and for control subjects 2 cells/mm; the coeliac patients had the same density as the patients with dermatitis herpetiformis (45 cells/mm). The higher density of cells bearing the alpha/delta T cell receptor in the epithelium (median 77 cells/mm) of untreated patients with dermatitis herpetiformis was associated with a gluten containing diet; in specimens taken from patients with dermatitis herpetiformis on a gluten free diet the median density was similar to that in the control subjects (44 v 39 cells/mm). The increase in the number of lymphocytes bearing the T cell receptor gamma/delta, particularly in the epithelium of the jejunum, seems to be a constant marker for these closely related diseases, whereas the density of alpha/delta+ T cells is dependent on the diet.     

Small bowel enema in non-responsive coeliac disease.

A small bowel enema was performed in patients with non-responsive coeliac disease, in coeliac patients on a normal diet (untreated) and those who had shown a good response to a gluten free diet, and in control subjects to determine whether there were any specific radiological features of the non-responsive state. A significant reduction in the average number of jejunal folds and an increase in the number of ileal folds (reversal of the jejunoileal fold pattern) was found in eight of nine non-responsive coeliac patients, one of seven untreated coeliac patients, and in none of the good responders or control subjects. This pattern identifies coeliac patients with a poor response to a gluten free diet who are likely to suffer major complications.     

Mucosal Enzyme Activity as a Quantitative Index of Early Functional Improvement in the Management of Coeliac Disease and Other Small Intestinal Diseases

Summary: The levels of alkaline phosphatase and the three disaccharidases, sucrase, lactase and maltase, have been measured in biopsy specimens from jejunal mucosa. Patients with small intestinal diseases have been compared with a control population consisting of patients and volunteer subjects. This quantitative index of mucosal function has been used to assess the early response to specific therapy of small intestinal diseases, including coeliac disease, giardiasis and Whipple's disease.
Biopsies from 13 patients with untreated coeliac disease showed a marked reduction of all enzymes studied. In most of these patients the results were well outside the normal range. Lactase concentration was most severely reduced. In eleven patients studied following commencement of a gluten free diet, mucosal enzyme levels were significantly higher and frequently within the normal range. Lactase levels, although higher than in untreated patients, usually remained below normal values.
In eight patients with coeliac disease, mucosal biopsies were compared before and after commencement of a gluten free diet. The repeat biopsies were taken as early as six weeks following commencement of treatment. In all patients mucosal enzyme levels were significantly increased although histological changes were frequently difficult to recognize.     

Ultrastructural analysis of plasma cells in coeliac patients.

By accurate morphometry the length of the rough endoplasmic reticulum (RER) was estimated in electron micrographs of a population of jejunal plasma cells in biopsies obtained from normal volunteers, untreated coeliac disease patients, and coeliac disease patients treated for one year on a gluten free diet. Increase in length or hypertrophy of the endoplasmic reticulum is an indication of increased protein synthesis of the cell (Ghadially, 1977). An increase compared with normal length was demonstrated in the mean length of the plasma cell rough endoplasmic reticulum in plasma cells obtained from both groups of coeliac patients. This estimate of increased protein production--in this case immunoglobulin--indicates an increase in the immunological activity of plasma cells even after treatment and suggests not only a reaction to dietary gluten but probably to other antigens also.     

Endomysial antibody: is it the best screening test for coeliac disease?

The sensitivities and specificities of the IgA and IgG antigliadin antibody and the IgA antireticulin antibody have been compared with the recently described endomysial antibody directed against the basement membrane of smooth muscle in monkey oesophagus. One hundred and seventeen patients with adult coeliac disease (21 untreated), 84 patients with inflammatory bowel disease, systemic lupus erythematosus and rheumatoid arthritis (comprising the disease control group), 47 normal controls and a miscellaneous group of 29 patients, who were selected because of a positive reticulin staining pattern, were investigated. These results were correlated with the degree of abnormality of the intestinal mucosa in patients with adult coeliac disease. Endomysial antibodies were found in all patients with untreated coeliac disease and subtotal villous atrophy and in 47% of patients on a non-strict gluten free diet. One patient on a strict gluten free diet was positive and had partial villous atrophy while all patients in disease control groups were negative. Results were variable with the antireticulin and antigliadin antibodies. Sensitivity and correlation with subtotal villous atrophy in the untreated patients was 100%. It is concluded that the endomysial antibody is superior to other current antibody tests and should be used in preference for the diagnosis of coeliac disease.     

Raised number of jejunal IgG2-producing cells in untreated adult coeliac disease compared with food allergy

The subclass distribution of IgG-producing immunocytes was studied by two colour immunohistochemistry with monoclonal antibodies in jejunal biopsy specimens from 10 adults with untreated coeliac disease, 11 coeliac disease patients on a gluten free diet, and seven patients with established food allergy. Paired immunofluorescence staining was performed with subclass specific murine monoclonal antibodies in combination with polyclonal rabbit antibody reagent to total IgG; the proportion of cells belonging to each subclass could thereby be determined. The ratio of IgG2 immunocytes was significantly higher (p less than 0.05) in untreated coeliac disease patients (median, 35.2%; range, 26.7-65.2%) than in those on a gluten free diet (median, 7.3%; range, 0-31.9%) or those having food allergy (median, 12.5%; range, 0-36.5%). The disparity in the local IgG2 response between patients with untreated coeliac disease and those with food allergy might be due to differences in the nature of the antigenic stimuli, dissimilar genetic 'make-up' of the subjects, or both.     

The cellular infiltrate of the jejunum in adult coeliac disease and dermatitis herpetiformis following the reintroduction of dietary gluten.

The cellular infiltrate of the jejunal mucosa has been studied in patients with both treated and untreated adult coeliac disease and dermatitis herpetiformis and serially in treated patients before and after the reintroduction of gluten to the diet. In adult coeliac disease and dermatitis herpetiformis the jejunal mucosa showed similar abnormalities of the cellular infiltrate which was characterized by an increase in intraepithelial lymphocytes and lamina propria plasma cells and eosinophils, with the greatest numbers of cells occurring in untreated patients. At 24-48 hours following a single 25-g gluten challenge there was an increase in lamina propria plasma cells, lymphocytes and eosinophils and intraepithelial lymphocytes. This rise was sustained after seven days on a gluten-containing diet for all of these cell groups except lamina propria lymphocytes. These responses were essentially similar in both adult coeliac disease and in those dermatitis herpetiformis patients who had jejunal lesions before treatment. In dermatitis herpetiformis patients with normal jejunal morphology on a normal diet there was an upward trend in lamina propria plasma cells and intraepithelial lymphocytes within one to three weeks of taking extra dietary gluten. These results are compatible with the view that more than one immunological mechanism may be responsible for the pathogenesis of the jejunal lesion of coeliac disease and dermatitis herpetiformis.     

Blisters in the small intestinal mucosa of coeliac patients contain T cells positive for cyclooxygenase 2.

BACKGROUND AND AIMS: Coeliac disease is characterised by atrophy of the villi and hyperplasia of the crypts in the mucosa of the small intestine. It is caused by an environmental trigger, cereal gluten, which induces infiltration of the mucosa by inflammatory cells. We hypothesised that these inflammatory cells express cyclooxygenase 2 (COX-2), an enzyme that contributes to the synthesis of pro and anti-inflammatory prostaglandins and is known to be expressed at sites of inflammation in the stomach and colon. We have investigated expression of COX-2 in the coeliac disease affected small intestinal mucosa where it may be an indicator of either disease induction or mucosal restoration processes. PATIENTS AND METHODS: Small intestinal biopsy samples from 15 coeliac patients and 15 non-coeliac individuals were stained immunohistochemically for COX-2. Samples from 10 of the patients were also stained after these patients had been on a gluten free diet for 6-24 months. Various cell type marker antigens were used for immunohistochemical identification of the type of cell that expressed COX-2. To further verify colocalisation of the cell type marker and COX-2, double immunoperoxidase and immunofluorescence methods were employed. Immunoelectron microscopy was used to investigate the subcellular location of COX-2. RESULTS: In all samples taken from coeliac patients, clusters of cells with strong immunoreactivity for COX-2 were found in those areas of the lamina propria where the epithelium seemed to blister or was totally detached from the basement membrane. These clusters were reduced in number or totally absent in samples taken after a gluten free diet. No such clusters were seen in any control samples. The density of COX-2 positive cells lining the differentiated epithelium decreased significantly from 13.5 (5.1) cells/10(5) microm(2) (mean (SD)) in the untreated patient samples to 6.5 (2.0) cells/10(5) microm(2) after a gluten free diet (p<0.001), and was 3.3 (1.9) cells/10(5) microm(2) in control samples (p<0.001 compared with untreated or diet treated coeliac samples). Staining for COX-2 was localised to CD3+ T cells and CD68+ macrophages in the mucosal lesions but not all of these cells were positive for COX-2. Immunoelectron microscopy revealed that the ultrastructure of the COX-2 positive cells resembled that of lymphocytes, and the immunoreaction was localised to the rough endoplasmic reticulum and the nuclear envelope. CONCLUSIONS: Our results show that in coeliac disease, blistering of small intestinal epithelial cells is associated with accumulation of COX-2 positive T cells, and the number of these cells decreases after a gluten free diet. These observations suggest that COX-2 mediated prostanoid synthesis contributes to healing of the coeliac mucosa and may be involved in maintenance of intestinal integrity.     

Low circulating insulin-like growth factor I in coeliac disease and its relation to bone mineral density.

BACKGROUND: Patients with coeliac disease have low bone mineral density (BMD), but the underlying mechanisms are unclear. Our aim was to study circulating insulin-like growth factor I (IGF-I) and its possible relationship to BMD in adults with untreated coeliac disease and after 1 year on a gluten-free diet. METHODS: In 29 consecutive adult coeliac patients fasting IGF-I and BMD (n = 28) were examined before and 1 year after starting a gluten-free diet. Intact parathyroid hormone (PTH) was measured (n = 20) before the gluten-free diet was started. RESULTS: Untreated coeliac patients had lower IGF-I values than controls matched for age and sex, and their BMD was low. A relationship was observed between BMD and IGF-I but not independent of age and body mass index. During the 1st year on a gluten-free diet BMD increased (P < 0.001), as did the circulating IGF-I levels in 21 of the 29 patients (P = 0.078). In the subgroup of 14 patients with normal initial PTH the increase in IGF-I correlated positively with the increase in BMD (femoral trochanter, r = 0.62, P < 0.05, and lumbar spine, r = 0.70, P < 0.02). CONCLUSIONS: BMD and circulating IGF-I levels are low in adults with untreated coeliac disease. In patients with normal initial PTH level there is an association between the change in BMD and circulating IGF-I, although this parallel increase may not be causally connected.     

Cell-mediated immunity to gluten within the small intestinal mucosa in coeliac disease.

Jejunal biopsies from controls and coeliac patients were maintained in organ culture in the presence of gluten fraction III. The culture media were assayed for evidence of lymphokine activity in a migration inhibition test using normal peripheral blood leucocytes. Significant inhibition of migration was produced by media from untreated coeliac patients compared with controls (P less than 0.005) or treated coeliac patients (P less than 0.001), indicating the production of a leucocyte migration inhibition factor (LIF) by untreated coeliac mucosa in response to gluten fraction III. The degree of inhibition correlated with the preculture interepithelial lymphocyte count in the coeliac biopsies (P less than 0.02). In six coeliac patients studied when on a normal diet and on a gluten-free diet, LIF was produced while on a normal diet, but not while on a gluten-free diet. These results suggest that a local cell-mediated immune reaction to gluten is present in the mucosa of patients with untreated coeliac disease but that this is reversed by treatment with a gluten-free diet.     

Intestinal permeability in coeliac disease: the response to gluten withdrawal and single-dose gluten challenge.

Intestinal permeability has been studied in 21 patients with coeliac disease in relapse and after gluten withdrawal using an oral test of intestinal permeability based on the simultaneous oral administration of two probe molecules. The increased absorption of the larger molecule (cellobiose) and the decreased absorption of the smaller (mannitol) found in untreated coeliac disease both returned to normal within five months of starting treatment, the abnormality in cellobiose absorption correcting more rapidly than that of mannitol. After exposure to a single oral dose of gluten, the intestinal permeability of six patients with treated coeliac disease became transiently abnormal with an increased absorption of cellobiose, returning to normal within one week. The possible structural and functional implications of these findings are discussed. The cellobiose/mannitol ratio appears to be of value in assessing the response to gluten withdrawal in coeliac disease, and also in monitoring patients who are already established on a gluten free diet by detecting dietary lapses and 'non-responding coeliac disease'. It may also offer an alternative to jejunal biopsy in patients subjected to gluten challenge.