Content uploaded by Carolina Ciacci
Author content
All content in this area was uploaded by Carolina Ciacci on Apr 24, 2018
Content may be subject to copyright.
Prevalence of Thyroid Disorders in Untreated Adult
Celiac Disease Patients and Effect of Gluten
Withdrawal: An Italian Multicenter Study
C. Sategna-Guidetti, M.D., Ph.D., U. Volta, M.D., C. Ciacci, M.D., P. Usai, A. Carlino,
L. De Franceschi, M.D., A. Camera, M.D., A. Pelli, M.D., Ph.D., and C. Brossa, Ph.D.
University of Torino, Torino; University of Bologna, Bologna; University of Napoli, Napoli; University of
Cagliari, Cagliari; and University of Perugia, Perugia, Italy
OBJECTIVES: Many afflictions have been associated with
celiac disease, but chance associations may exists. The aim
of this study was to establish, by means of a multicenter
prospective study, the prevalence of thyroid impairment
among adult patients with newly diagnosed celiac disease
and to evaluate the effect of a 1-yr gluten withdrawal on
thyroid function.
METHODS: A total of 241 consecutive untreated patients and
212 controls were enrolled. In 128 subjects a thorough
assessment, including intestinal biopsy, was repeated within
1 yr of dietary treatment. Thyroid function was assayed by
measuring the levels of TSH, free T3, free T4, thyroperoxi-
dase, and thyroid microsome antibodies.
RESULTS: Thyroid disease was 3-fold higher in patients than
in controls (p⬍0.0005). Hypothyroidism, diagnosed in 31
patients (12.9%) and nine controls (4.2%), was subclinical
in 29 patients and of nonautoimmune origin in 21. There
was no difference regarding hyperthyroidism, whereas au-
toimmune thyroid disease with euthyroidism was present in
39 patients (16.2%) and eight controls (3.8%). In most
patients who strictly followed a 1-yr gluten withdrawal (as
confirmed by intestinal mucosa recovery), there was a nor-
malization of subclinical hypothyroidism. Twenty-five per-
cent of patients with euthyroid autoimmune disease shifted
toward either a subclinical hyperthyroidism or subclinical
hypothyroidism; in these subjects, dietary compliance was
poor. In addition, 5.5% of patients whose thyroid function
was normal while untreated developed some degree of thy-
roid dysfunction 1 yr later.
CONCLUSIONS: The greater frequency of thyroid disease
among celiac disease patients justifies a thyroid functional
assessment. In distinct cases, gluten withdrawal may single-
handedly reverse the abnormality. (Am J Gastroenterol 2001;
96:751–757. © 2001 by Am. Coll. of Gastroenterology)
INTRODUCTION
Many afflictions have been associated with celiac disease
(CD) (1); however, because the high prevalence of this
disease (2), chance associations may exist. Recent studies
have shown an increased incidence of thyroid diseases (TD)
among patients with CD (3, 4) but, thus far, data on the
prevalence and clinical significance of this association are
lacking.
This article describes a multicenter study, carried out
under the auspices of the Italian “Club del Tenue,” that
aimed to establish the prevalence of thyroid involvement
among adult patients newly diagnosed with CD and to
evaluate the effect of a 1-yr gluten withdrawal on thyroid
function.
MATERIALS AND METHODS
Five adult study centers from various regions of Italy (Fig.
1) took part in this prospective study.
Patients
Between January 1996 and July 1998, 241 consecutive adult
patients who were newly diagnosed with CD were enrolled.
The study population included 64 men (median age 29 yr,
range 18–82 yr) and 177 women (median age 31 yr, range
19–85 yr).
The diagnosis of CD was made according to the accepted
criteria (5), i.e., clinical history and typical histological
appearance of the small intestinal mucosa. In 224 of 241
subjects (93%), CD-related serology, i.e., antiendomysium
antibodies (IgA-EmA), was also positive. Persistence of
symptoms before diagnosis ranged from 6 to 120 months in
161 patients (66.8%), whereas symptoms had been present
in 41 patients (17%) since childhood. The clinical pattern
was typical in 40%, atypical in 44%, and silent in 16% of
patients, respectively (6); in the silent disease group, diag-
nosis was reached because of a screening program in a
family study.
Control Subjects
Because an Italian survey to appraise the prevalence of TD
(both subclinical or overt) in the general population had not
yet, to our knowledge, been performed, sera from 212 con-
trol subjects matched for sex, ethnic origin, and age (⫾3 yr)
THE AMERICAN JOURNAL OF GASTROENTEROLOGY Vol. 96, No. 3, 2001
© 2001 by Am. Coll. of Gastroenterology ISSN 0002-9270/01/$20.00
Published by Elsevier Science Inc. PII S0002-9270(00)02410-2
were screened for TD as well. Controls (64 men and 148
women) were healthy volunteers recruited among medical
and nursing staff, blood donors, or patients affected by
chronic obstructive pulmonary disease (COPD), peptic ul-
cer, or nonulcer dyspepsia (NUD) attending for upper en-
doscopy; none had a past or present history of thyroid
function impairment. Exclusion criteria included age ⬍19 yr
and the presence of disease that may affect thyroid function
(i.e., chronic liver or renal disease, malignancy, malnutri-
tion) or the use of medication known to influence serum
TSH or fT4 (i.e., dopamine, glucocorticoids, heparin, furo-
semide). In all participants, CD was excluded by means of
clinical findings and an IgA-EmA test or by duodenal biopsy
performed during endoscopy.
Study Design
At the time of entry into the study, all patients were un-
treated. Initial data collection included clinical history (6)
and findings, associated diseases, clinical and biochemical
nutritional assessment, intestinal biopsy, and thyroid func-
tion appraisal: thereafter, patients were instructed to follow
a gluten-free-diet (GFD) and were considered as treated.
Patients with overt hypo- or hyperthyroidism were started
on an appropriate medical treatment. A thorough assess-
ment, including intestinal biopsy, was repeated within 12 ⫾
1 months.
The study was performed according to the principles of
the Declaration of Helsinki, and informed consent was ob-
tained from each patient. The research was approved by the
Scientific Committee of the “Club del Tenue.”
Nutritional Assessment
Nutritional status was evaluated by calculating the body
mass index (BMI; kg/m
2
, reference range: men ⬎20,
women ⬎19). Biochemical assessment, performed by stan-
dard methods, included the following: Hb (reference range
14–18 g/dl (for men) and 12–16 g/dl for women); serum
albumin (Alb, reference range 3.6–5.0 g/dl); serum iron (Fe,
reference range 59–158
g/dl for men and 37–145
g/dl for
women).
Thyroid Function Assessment
Serum fT3 and fT4 were estimated by radioimmunoassay
(Ortho-Clinical Diagnostic, Amersham, UK). The sensitiv-
ities of the assays were 0.768 pmol/L for fT3 and 0.769
pmol/L for fT4, respectively, with intra-assay percentage
coefficients of variation (CV) of 4% and 3.8%, and inter-
assay CV of 6.4% and 6.2%.
Serum TSH levels were evaluated by immunoradiometric
assay (Biocode, Lie`ge, Belgium) on solid phase. The min-
imum measurable concentration was at 0.05
IU/ml. The
interassay CVs were 5% and 5.5%, respectively.
Serum thyroid microsome antibodies (TM-Ab) were ap-
praised by immunoradiometric assay (Biocode). The detec-
tion limit has been evaluated at 2.5 IU/ml. The intra-assay
and interassay CVs were 8.1% and 7.9%, respectively. Se-
rum levels of thyroperoxidase antibodies (TPO-Ab) were
detected by radioimmunoassay (Biocode). The sensitivity
has been evaluated at 7 IU/ml. The intra and interassay CVs
were 8.9% and 9.4%, respectively.
We considered as normal the following serum levels:
TSH 0.3–3.0
IU/ml, fT3 2.2–4.7 ng/L, fT4 8–20 ng/L,
TM-Ab ⱕ50IU/ml, and TPO-Ab ⱕ50 IU/ml.
Thyroid disease was classified, according to the American
Thyroid Association guidelines, as follows (7): overt hyper-
thyroidism: TSH ⬍0.3
IU/ml and fT4 ⬎20 ng/ml; sub-
clinical hyperthyroidism: TSH ⬍0.3
IU/ml and fT4 in the
normal range; overt hypothyroidism: TSH ⬎3
IU/ml and
fT4 ⬍8 ng/ml; subclinical hypothyroidism: TSH ⬎3
IU/ml and a normal fT4; subclinical autoimmune (AI)
hypothyroidism: TSH ⬎3
IU/ml and a normal fT4 and
⬎100 IU/ml serum levels of TPO-Ab or TM-Ab or both;
subclinical nonautoimmune hypothyroidism (NAI): TSH
⬎3
IU/ml and a normal fT4, without autoantibody serum
level increase; and autoimmune thyroid disease with euthy-
roidism (ATDE): TSH and fT4 in the normal ranges and
⬎100 IU/ml serum levels of TPO-Ab or TM-Ab or both. All
reference limits were provided by the manufacturers, and all
biochemical values were compared with those for the indi-
vidual analyte reference range established in the laboratory.
Intestinal Morphology
All patients underwent intestinal biopsy from the second
part of the duodenum during upper endoscopy. Mucosal
tissue was graded according to the Marsh criteria (8).
Figure 1. Centers participating in the study.
752 Sategna-Guidetti et al. AJG – Vol. 96, No. 3, 2001
Statistical Analysis
Statistical data were generated using the MedCalc software
for Windows (Mariakerke, Belgium). To ascertain whether
possible differences in TD between patients and controls
may be clinically meaningful and worthwhile to detecting,
we previously estimated the required sample size by calcu-
lating the probabilities
␣
and

of a type I and type II error,
respectively.We calculated that 184 patients and 184 con-
trols should be enrolled by assuming that the expected
prevalence of TD among the general population (4) (and
therefore among our controls) is 5%, that the relative risk for
TD in CD is 3-fold higher and by accepting an
␣
error (type
I error) of 5% and a power (1-

)⫽90%.
Data were analyzed using the
2
test (or Fisher’s exact
test, where indicated) for differences in frequencies, the t
test (paired or unpaired, as appropriate) for comparison of
means, and the r correlation coefficient. For all statistical
analysis, a two-tailed pvalue ⬍0.05 was considered sig-
nificant.
RESULTS
Thyroid Function and Clinical Findings
at Diagnosis (Untreated Subjects)
As shown in Table 1, overall TD was found in 73 (61
women and 12 men) of 241 (30.3%) patients with CD, and
in 24 (19 women and five men) of 212 controls (11.3%)
(
2
⫽23.004, p⬍0.0005). However, the differences were
statistically significant only for women (
2
⫽19.165, p⬍
0.0005).
Overall hypothyroidism, which had been diagnosed in 31
patients and nine controls (p⬍0.003), was subclinical in 29
patients and eight controls and was overt in the remainder;
however, the differences were statistically significant only
for women (
2
⫽8.284, p⫽0.0045). In 21 patients and four
controls, a NAI hypothyroidism was found, whereas in 10
patients and five controls an AI hypothyroidism was present.
NAI hypothyroidism was significantly higher (
2
⫽11.821,
p⫽0.0006) than AI.
Table 1. Prevalence of Thyroid Function Impairment Among Newly Diagnosed Celiac Disease Patients and Controls
Thyroid Function
Untreated
Celiac Disease
Patients Controls
Hypothyroidism
Number (%) 31/241 (12.9%) 9/212 (4.2%)
2
⫽9.362, p⬍0.003
Sex 26F,5M 7F,2M
Hyperthyroidism
Number (%) 3/241 (1.2%) 7/212 (3.3%) NS
Sex 2F,1M 7F,0M
Autoimmune thyroid disease
with euthyroidism
Number (%) 39/241 (16.2%) 8/212 (3.8%)
2
⫽17.366, p⬍0.0005
Sex 33F,6M 5F,3M
Total
Number (%) 73/241 (30.3%) 24/212 (11.3%)
2
⫽23.004, p⬍0.0005
Sex 61 F, 12 M 19 F, 5 M
Table 2. Nutritional Assessment in Untreated Patients With Celiac Disease With and Without Associated Thyroid Function
Impairment
Characteristic
Celiac Disease
Without
Thyroid
Disease Celiac Disease and
Thyroid Disease pValue*
BMI kg/m
2
M 20.6, F 20.4 M 22.3, F 19.7
M⬍20 kg/m
2
21/52 (40.4%) 1/12 (8.33%) 0.0448 (Fischer’s),
F⬍19 kg/m
2
37/116 (31.9%) 22/61 (36%) NS
Hb g/dl M 12.8, F 11.4 M 12.1, F 11.6
M⬍14 g/dl 19/52 (36.5%) 8/12 (66.7%) NS
F⬍12 g/dl 58/119 (50%) 26/61 (42.6%) NS
Serum albumin g/dl 4.7 4.4
⬍3.5 g/dl 18/168 (10.7%) 5/73 (6.8%) NS
Serum iron
g/dl M 80, F 48 M 63, F 51
M⬍59
g/dl 11/52 (21.1%) 6/12 (50%) NS
F⬍37
g/dl 45/116 (38.8%) 16/61 (26.2%) NS
Data are median values and number of patients with values below the lower limit. BMI ⫽body mass index; F ⫽female; M ⫽male.
*pValues determined by
2
test unless stated otherwise.
753AJG – March, 2001 Thyroid Disorders in Untreated Adult Celiac Disease
Hyperthyroidism was diagnosed in three patients and
seven controls, and was subclinical in two patients and in
five controls.
ATDE was present in 39 patients and eight controls; but,
again, the difference was statistically significant only in
women (
2
⫽16.744, p⬍0.0005).
The presence of TD was not influenced by geographic
factors. The age of patients and controls with associated TD
did not differ from that of patients and controls with a
normal thyroid function (median age 33 and 32.5 yr, and 29
and 30.5 yr, respectively). There was no variation in TD
prevalence in age groups. The clinical pattern of CD (clas-
sic, atypical, silent) (6), presenting symptoms (diarrhea,
anaemia, weight loss, dyspepsia, bloating, fatigue, weak-
ness, osteoporosis), associated diseases (dermatitis herpeti-
formis, insulin-dependent diabetes mellitus, arthritis, viti-
ligo, aphthous stomatitis), and histological scores (8) were
comparable in patients with and without TD. In three of 73
patients (0.96%) with thyroid function impairment, TD di-
agnosis preceded CD recognition.
Nutritional Assessment
At diagnosis, median BMI values and Hb, serum iron, and
albumin levels were similar in patients with and without TD.
In each group, a number of subjects showed abnormal values
that did not reach a statistical significance, with the exception
of BMI among men (Table 2). No significant correlation was
found among nutritional indices and TSH or fT3 values.
Table 3. Thyroid Function Outcome in 128 Celiac Disease Patients Within 1 Yr of Gluten Withdrawal
Thyroid Function Assessment
at Diagnosis Number of Patients Reassessed Outcome Within 1 Yr
1 overt 3drug therapy 31 subclinical AI hypothyrodism
Autoimmune (AI) 6
33 ATDE
5 subclinical
32 unchanged
Hypothyroidism
310/14 recovery (71.4%)
Nonautoimmune (NAI) 14 subclinical 33 no change
31 subclinical AI
Hyperthyroidism 1 overt 3drug therapy 31 subclinical hyperthyroidism
33 normalization
Autoimmune thyroid disease with 39 unchanged
euthyroidism (ATDE) 16
33 subclinical AI hypothyroidism
31 subclinical hyperthyroidism
32 NAI hypothyroidism
Normal 91 31 subclinical hyperthyroidism
32 ATDE
Subclinical hypothyroidism normalized in 10 of 14 patients with NAI, seven of whom had adhered strictly to the diet. In subjects without thyroid function improvement,
compliance with diet was generally poor. In five patients with a normal thyroid function at diagnosis, there was a shift toward some impairment; only one of these did not strictly
comply with the diet.
AI/NAI ⫽autoimmune/nonautoimmune thyroid disease; ATDE ⫽autoimmune thyroid disease with euthyroidism.
4/16 (25%)
5/91 (5.5%)
754 Sategna-Guidetti et al. AJG – Vol. 96, No. 3, 2001
Effect of a 1-Yr Gluten-Free Diet:
Thyroid Function and Clinical Findings
A total of 128 patients (53.1%) underwent a complete re-
assessment within 12 ⫾1 months after the beginning of
gluten withdrawal, 91 with a normal thyroid function and 37
with an impaired function at diagnosis. An exhaustive re-
view of diet to assess compliance did not show a significant
difference between the two groups. Intestinal biopsy re-
peated in 75 of 128 (58.6%) subjects showed a mucosal
recovery in 43 (57.3%) and persisting partial or total atrophy
in the remaining 32 (42.7%).
The outcome of the 128 patients’ thyroid function is
shown in Table 3. Three points must be stressed. First,
subclinical hypothyroidism normalized in 10 of 14 (71.4%)
patients with NAI disease, seven of whom had presumably
adhered strictly to the GFD; in three of five (60%) patients
with AI disease (two of whom were compliant with the
diet), there was a shift to ATDE. In four of five subjects with
no improvement in thyroid function, compliance with the
diet was poor, as demonstrated by the incomplete recovery
of the intestinal mucosa at duodenal biopsy. Second, out of
four of 16 (25%) patients with ATDE, one patient developed
a subclinical hyperthyroidism, whereas three had AI subclini-
cal hypothyroidism; three of these subjects did not strictly
comply with the diet. Third, 1 yr later, out of five of 91 (5.5%)
patients with a normal thyroid function at diagnosis, one patient
had developed subclinical hyperthyroidism, two ATDE, and
two NAI subclinical hypothyroidism; only one of these patients
did not strictly comply with the diet.
Nutritional Assessment
A 1-yr gluten withdrawal led to a significant improvement
of nutritional indices (Table 4), which was comparable in
patients with and without TD, as shown in Figure 2. In most
patients with persistent mucosal impairment, nutritional in-
dices continued to be abnormal.
DISCUSSION
Our study confirms that patients with celiac disease are at
risk for developing thyroid disease, with an overall 3-fold
higher frequency than in controls. In both populations,
women were significantly more affected than men; preva-
lence of TD did not differ between age groups. These data
should be representative of the Italian population overall, as
a large number of subjects, evenly distributed across the
country, were examined.
In seven patients, diagnosis of TD preceded recognition
of CD (i.e., the disorder developed in a condition of unrec-
ognized and untreated CD).
Clinical presentation, age at diagnosis and incidence of
associated diseases were comparable in patients with and
without TD: our data do not seem to support, at least as far
as thyroid function involvement is concerned, the theory of
Ventura et al. (9) that age at diagnosis, which indirectly
mirrors the duration of gluten exposure, may increase the
risk of developing autoimmune disorders. The discrepancy
could be due to a bias caused by the age of our patients, i.e.,
adults, versus adolescents in the study by Ventura et al.
Our findings confirm the infrequent impairment of nutri-
tional status in untreated CD both with and without TD,
although approximately one-half had anemia and one-third
had iron deficiency. Notwithstanding a sufficient nutritional
status, it is probable that in 21 of 31 patients hypothyroidism
was not of autoimmune origin but rather was attributable to
a decreased thyroid hormone synthesis, induced either by an
iodine organification defect or by a functional hypothalam-
ic–pituitary disturbance consequent to isolated malnutrition
(10, 11). This hypothesis may be supported both by the
relationship between normalization of thyroid function, by
means of gluten withdrawal alone, and by the histological
recovery of duodenal mucosa, or, conversely, by the ab-
sence of thyroid function improvement in subjects with
persistent villous atrophy. This strengthens the view that
gluten withdrawal by itself is able to eliminate the main
etiological factor of NAI hypothyroidism.
The prevalence of thyroid antibody positivity in euthyroid
subjects was 4-fold higher in CD than in controls. It has
been postulated that thyroid autoimmunity is enhanced in
iodine-depleted geographical areas, where an increased
prevalence of goiter and hyperthyroidism of nonautoim-
Table 4. Nutritional Outcome in Celiac Disease Patients
Characteristic Untreated Treated pValue
BMI kg/m
2
M 21.15 ⫾3.2, F 22.6 ⫾3.3 M 20.8 ⫾3.68, F 23 ⫾3.44 0.043 (M)*, NS (F)*
M⬍20 kg/m
2
22/64 (34.4%) 7/31 (22.6%) NS
F⬍19 kg/m
2
59/177 (33.33%) 13/97 (13.4%) 0.0006
Hb g/dl 12.74 ⫾6.39 13.6 ⫾1.64 NS*
M⬍14 g/dl 27/64 (42.2%) 4/31 (12.9%) 0.0091
F⬍12 g/dl 84/177 (47.5%) 10/97 (10.3%) ⬎0.0005
Serum albumin g/dl 4.9 ⫾0.503 5.13 ⫾0.57 ⬍0.0001*
⬍3.5 g/dl 23/241 (9.54%) 5/128 (3.9%) NS
Serum iron
g/dl 63 ⫾38.13 92 ⫾33.7 ⬍0.0001*
M⬍59
g/dl 17/64 (26.6%) 0.31 (0%) 0.0044
F⬍36
g/dl 61/177 (34.4%) 2/97 (2.04%) ⬍0.0005
Data given as mean ⫾SD and number of patients with values below the lower limit in each group. BMI ⫽body mass index; F ⫽female; M ⫽male.
* Statistical analysis by paired Student ttest; other statistical analysis by unpaired student tand
2
test.
755AJG – March, 2001 Thyroid Disorders in Untreated Adult Celiac Disease
mune origin are more frequent (12). Although thyroid ul-
trasonography was not performed because of a possible
operator-dependent bias (as this was a multicenter study),
this statement does not apply to our patients; the rate of
antibody positivity was similar in regions with variable
iodine supplies (Napoli and Cagliari are located by the sea),
and both patients and controls had a low overall hyperthy-
roidism rate. Although the clinical significance of these
antibodies in CD is still unclear, there is probably a higher
propensity for thyroid autoimmunity; proposed mechanisms
of autoimmune endocrine disease involve a sequence of
immune, inflammatory events in a genetically susceptible
individual. In most cases, the immune response to the target
cell progressively destroys the endocrine gland, and hypo-
function is the main clinical manifestation (13). This mech-
anism seems to be sustained by the fact that, 1 yr later, three
out of 16 (19%) patients with ATDE developed AI subclin-
ical hypothyroidism and one (6%) a subclinical hyperthy-
roidism. Thus, a longitudinal follow-up would seem neces-
sary in patients who have a positive AI thyroid serology but
who are currently euthyroid; this course would help to
evaluate whether these findings should be considered only
as an epiphenomenon or should be of prognostic help. This
circumstance could change the attitude toward the cost-
effectiveness of screening, as it could be very useful to have
some advance knowledge about subjects who may become
clinically hypothyroid (or, less likely, hyperthyroid).
The importance of gluten withdrawal was further under-
lined, 1 yr later, by the normalization of antithyroid anti-
bodies in three of these 16 ATDE patients (19%), who were
strictly diet compliant, although there exists the possibility
of a spontaneous fluctuation of antibodies.
Reprint requests and correspondence: Prof. Carla Sategna-
Guidetti, Cattedra di Gastroenterologia, Dipartimento di Medicina
Interna, Corso AM Dogliotti, 14. 10126 Torino, Italy.
Received May 9, 2000; accepted Sep. 14, 2000.
REFERENCES
1. Cooke WT, Holmes GKT. Coeliac disease, and associated
disorders. In: Coeliac disease. Cooke WT, Holmes GKT, eds.
London: Churchill Livingstone, 1984:225–48.
2. Catassi C. Screening of coeliac disease. In: Maki M, Collin P,
Visakorpi JK, eds. Coeliac disease. Proceedings of the 7th
International Symposium on coeliac disease. Tampere,
Finland: Coeliac Disease Study Group, 1997:23–34.
3. Counsell CE, Taha A, Ruddell WSJ. Coeliac disease, and
autoimmune thyroid disease. Gut 1994;35:844–6.
4. Sategna-Guidetti C, Bruno M, Mazza E, et al. Autoimmune
thyroid disease, and coeliac disease. Eur J Gastroenterol Hepa-
tol 1998;10:927–31.
5. Walker Smith JA, Guandalini S, Schmitz J, et al. Revised
criteria for diagnosis of coeliac disease. Arch Dis Child 1990;
65:909–11.
6. Ferguson A, Arranz E, O’Mahony S. Clinical, and patholog-
ical spectrum of coeliac disease—Active, silent, latent, poten-
tial. Gut 1993;34:150–1.
7. Surks MI, Chopra U, Mariash CN, et al. American Thyroid
Association guideline for use of laboratory tests in thyroid
disorders. JAMA 1990;263:1529–32.
Figure 2. Percentage of abnormal results, as compared to initial values (untreated) most of nutritional indices showed a significant
improvement (
2
test, on actual numbers) after a 1-yr gluten withdrawal (treated). No statistical difference was found between patients
with and without associated thyroid disease. BMI ⫽body mass index; s-albumin ⫽serum albumin; s-iron ⫽serum iron.
756 Sategna-Guidetti et al. AJG – Vol. 96, No. 3, 2001
8. Marsh MN. Morphology, and immunopathology of the jejunal
lesion in gluten-sensitivity. Eur J Gastroenterol Hepatol 1991;
3:108–14.
9. Ventura A, Magazzu G, Greco L, for the SIGEP Study Group
for Autoimmune Disorders in Celiac Disease. Duration of expo-
sure to gluten, and risk for autoimmune disorders in patients with
celiac disease. Gastroenterology 1999;117:297–303.
10. Carter JN, Eastman CJ, Corcoran JM, et al. Effect of severe,
chronic illness on thyroid function. Lancet 1974;ii:971–4.
11. Farthing MJG, Rees LR, Edwards CRW, et al. Thyroid hor-
mones and the regulation of thyroid function in men with
coeliac disease. Clin Endocrinol 1982;16:525–35.
12. Szabolcs I, Podoba J, Feldkamp J, et al. Comparative screen-
ing for thyroid disorders in old age in areas of iodine defi-
ciency, long term iodine prophylaxis, and abundant iodine
intake. Clin Endocrinol 1997;47:87–92.
13. Baker JR Jr. Autoimmune endocrine disease. JAMA 1997;
278:1931–7.
757AJG – March, 2001 Thyroid Disorders in Untreated Adult Celiac Disease