Serologic testing is useful in screening both persons with typical gastrointestinal manifestations of celiac disease and those without gastrointestinal symptoms, such as children with idiopathic short stature and dental anomalies [6-11]. Serologic testing for celiac disease in patients with insulin-dependent diabetes mellitus has been evaluated in Finland using antireticulin antibody testing [4, 12]. Collin and colleagues [12] found that 8 of 195 patients with insulin-dependent diabetes mellitus had celiac disease (observed prevalence of 1 per 25 patients); the expected prevalence is 1 per 1700 patients in the general population of Finland. If this observation can be confirmed among the ethnically diverse population of the United States, then recommending screening for celiac disease in patients with insulin-dependent diabetes mellitus may be reasonable.

The antiendomysial antibody test has been shown to be more sensitive than (90% to 100% compared with 68% to 91%) and as specific as (99% to 100% for both tests) the antireticulin assay when screening for celiac disease [13, 14]. By using the more sensitive antiendomysial antibody test to screen for celiac disease in patients with insulin-dependent diabetes mellitus, the true prevalence of coincident disease may be shown to be more frequent than previously identified.

We determined the prevalence of celiac disease in a cohort of North American patients with insulin-dependent diabetes mellitus by screening with the antiendomysial antibody test, and we described the clinical characteristics of patients with co-existent disease.

Methods

Top Methods Results Discussion Author & Article Info References

Study Design

All patients followed by the endocrinology service at our medical center were screened for insulin-dependent diabetes mellitus. In addition, the computerized pharmacy records of all patients for whom insulin had been prescribed in the last 10 years were reviewed to identify additional potential study candidates. Fifty consecutive candidates were identified, and histories were reviewed to ensure that each patient met American Diabetic Association criteria for insulin-dependent diabetes mellitus: age at disease onset younger than 20 years, history of diabetic ketosis, or C-peptide level less than 0.1 ng/mL [15]. All but 3 patients met the established criteria and had serologic screening for celiac disease with the antiendomysial antibody test. Patients with positive antibody test results had endoscopy with biopsy to evaluate for histologic evidence of celiac disease, multiphasic biochemical screening for nutritional deficiencies, serum D-xylose testing, and bone densitometry.

Antiendomysial Antibody Test

Slides of monkey esophagus (Scimedx Corporation, Danville, New Jersey) were incubated with 20 µL of patient serum samples diluted 1:10 in phosphate-buffered saline for 30 minutes at room temperature. Fluorescein isothiocyanate-conjugated antibody to human IgA was added to each well, and patient samples were incubated for an additional 30 minutes at room temperature. Fluorescent staining of smooth-muscle fiber endomysium at a titer of more than 1:10 was considered positive. Known positive and negative controls were used with each group of samples tested.

Histologic Examination

Patients with positive antibody test results had upper endoscopy; meperidine and diazepam were titrated to allow conscious sedation. Pinch mucosal biopsy specimens were obtained from the fourth portion of the duodenum and proximal jejunum. At least four biopsy specimens were obtained from each patient. Tissue was placed on filter paper with the luminal surface up before being immersed in formalin.

Criteria for Diagnosis of Celiac Disease

Celiac disease was diagnosed using internationally accepted criteria [16]: 1) positive antiendomysial antibody test results, and 2) histologic evidence of gluten-sensitive enteropathy (loss of villous architecture, crypt hyperplasia, and proliferation of intraepithelial lymphocytes) [17].

Bone Mineral Densitometry

Two patients with insulin-dependent diabetes mellitus and positive antibody test results had lumbar bone mineral densitometry by DXA (HOLOGIC QDR-1000, Lunar Corporation, Madison, Wisconsin). Measurements were compared with those of age-, sex-, and ethnicity-matched controls to estimate the variance in bone mineral density from normal. The bone mineral densitometry values are expressed as standard deviations (SD) from the control mean to estimate risk for bone fracture; for each SD below the control mean, the risk for bone fracture doubles [18].

Results

Top Methods Results Discussion Author & Article Info References

Forty-seven of the 50 patients screened (94%) met the American Diabetic Association criteria for insulin-dependent diabetes mellitus. Patients ranged in age from 22 years to 65 years (mean, 40 years). The ratio of men to women was 6:5. The group consisted of 38 white patients (82%), 7 Hispanic patients (14%), and 2 black patients (4%).

Three of the 47 patients with insulin-dependent diabetes mellitus (2 white men and 1 Hispanic woman) had positive antibody test results (Table 1). They ranged in age from 31 years to 45 years. The positive antiendomysial antibody titers ranged from 1:160 to 1:1280. Small-bowel biopsy specimens from all 3 patients showed complete villous atrophy, crypt hyperplasia, and lymphocyte infiltration (Figure 1).

View larger version (139K): [in this window] [in a new window]   Figure 1. Photoµgraph of small-bowel biopsy specimen from patient T9369. Shows characteristic histologic findings of celiac disease: marked villous atrophy, crypt hyperplasia, and expansion of the lamina propria with lymphocytes.

Bone mineral densitometry measurements for two of the three patients with serologic and histologic evidence of celiac disease were 0.8 and 1.1 SDs below the measurements of their age- and sex- matched controls. The third patient moved from our region of health care delivery, precluding bone densitometry testing.

Clinical signs and symptoms of malabsorption and maldigestion were not evident from directed questioning or routine biochemical testing. All patients were at or over their ideal body weights. Bowel habits were normal in consistency and frequency, except for those of patient P3471, who reported having 3 to 4 stools per day with a loose stool consistency.

Anemia was not evident in any of the patients tested. Absorption of D-xylose was abnormal in one of the two patients tested with bone mineral densitometry. None of the patients had skin rashes consistent with dermatitis herpetiformis.

Analysis

Three of 47 patients with insulin-dependent diabetes mellitus (6.4%; 95% CI, 1.4% to 17.5%) had positive antiendomysial antibody test results; all 3 had small-bowel biopsy specimens consistent with celiac disease. The CI lies entirely above the estimated prevalence of celiac disease expected in persons of the same age and sex in the general U.S. population, which ranges from 0.1% to 0.02%. Thus, celiac disease appears to be significantly more common in patients with insulin-dependent diabetes mellitus than in the general U.S. population (P < 0.001) [19].

Discussion

Top Methods Results Discussion Author & Article Info References

We identified 3 of 47 (6%) patients with insulin-dependent diabetes mellitus who had serologic and histologic evidence of gluten-sensitive enteropathy. This prevalence is similar to that previously reported by Collin and colleagues [12], who found that 4% of their patients with insulin-dependent diabetes mellitus had coexistent celiac disease. We had anticipated that screening for celiac disease might be lower in our study because of the diversity of our study sample (82% white patients, 14% Hispanic patients, and 4% black patients). The antiendomysial antibody test used to screen for celiac disease in our study is more sensitive than the antireticulin antibody test used by Collin and colleagues in their screening study [13, 14]. Use of the more sensitive test may be the reason why we found a higher occurrence of coexistent disease than we had anticipated.

In our study, clinical history and routine biochemical testing did not help to predict which diabetic patients had celiac disease, indicating that specific and sensitive serologic testing for antiendomysial antibody may be needed to detect subclinical disease in these patients. Absence of substantial biochemical deficiencies on routine analysis suggests that coexistence of this disease is not clinically important and that the commercial cost of the antiendomysial test ($50.00; Specialty Laboratories, Santa Monica, California) may be unjustified. However, bone densitometry measurements from the two patients we tested suggested that significant disturbances in calcium and bone homeostasis had occurred before routine biochemical tests showed abnormalities. Our findings are consistent with those of previous studies that identified patients with celiac disease to be at risk for subclinical reductions in bone mineral densitometry [20]. Longitudinal information on the effects of a gluten-free diet for patients with subclinical celiac disease is necessary before further recommendations are made for these patients, who already have intense diabetic dietary restrictions.

Celiac disease is often a latent clinical syndrome [8, 10, 11]. Nearly all patients who are identified as having latent celiac disease are positive for the HLA-DQ2 allele [6, 9]. The reason why some patients develop active celiac disease with complications is unknown, but loss of protective factors or exposure to aggressive factors (increased dietary gluten, gut hyperpermeability, gastrointestinal infections, non-HLA genes, nutritional deficiencies) may have some influence. It is unclear why none of our patients had overt symptoms of gluten-sensitive enteropathy. Although both of the patients tested showed diminished bone mineral density, only one of three patients had clinical symptoms suggestive of celiac disease (that is, altered bowel habits).

Our results suggest that the coincidence of gluten-sensitive enteropathy and insulin-dependent diabetes mellitus is relatively common (6%). In addition, clinical history and routine biochemical testing are insensitive in predicting which patients with insulin-dependent diabetes mellitus will have gluten-sensitive enteropathy. Clinicians should be aware of this association and should consider doing serologic testing for gluten-sensitive enteropathy in patients with insulin-dependent diabetes mellitus when symptoms or biochemical markers suggest malabsorption. Larger studies are needed to determine whether routine antiendomysial screening of all patients with insulin-dependent diabetes mellitus is warranted.

This work was presented at the 95th Annual Meeting of the American Gastroenterologic Association held in San Diego, California, 14-17 May 1995.

Dr. Merenich: 5487 South Idalia Way, Aurora, CO 80015.

Dr. Lieberman: Department of Clinical Investigation, Building 601, Fitzsimons Army Medical Center, Aurora, CO 80045.

Mr. Long: 2575 South Washington Street, Denver, CO 80210-5739.

Dr. Davis: 496 Glen Circle Drive, Maysville, KY 41056.

Dr. McNally: Gastroenterology Service, Department of Medicine, Fitzsimons Army Medical Center, Aurora, CO 80045-5000.