Mortality in patients with coeliac disease and their relatives: A cohort study

Article (PDF Available)inThe Lancet 358(9279):356-61 · September 2001with46 Reads
DOI: 10.1016/S0140-6736(01)05554-4 · Source: PubMed
Abstract
Although previous studies have shown increased mortality in patients with coeliac disease and their relatives, no data are available in relation to different patterns of clinical presentation. We assessed mortality in patients with coeliac disease and their first-degree relatives. We enrolled, in a prospective cohort study, 1072 adult patients with coeliac disease consecutively diagnosed in 11 gastroenterology units between 1962 and 1994, and their 3384 first-degree relatives. We compared the number of deaths up to 1998 with expected deaths and expressed the comparison as standardised mortality ratio (SMR) and relative survival ratio. 53 coeliac patients died compared with 25.9 expected deaths (SMR 2.0 [95% CI 1.5-2.7]). A significant excess of mortality was evident during the first 3 years after diagnosis of coeliac disease and in patients who presented with malabsorption symptoms (2.5 [1.8-3.4]), but not in those diagnosed because of minor symptoms (1.1 [0.5-2.2]) or because of antibody screening (1.2 [0.1-7.0]). SMR increased with increasing delay in diagnosis and for patients with poor compliance with gluten-free diet. Non-Hodgkin lymphoma was the main cause of death. No excess of deaths was recorded in relatives with coeliac disease. Prompt and strict dietary treatment decreases mortality in coeliac patients. Prospective studies are needed to clarify the progression of mild or symptomless coeliac disease and its relation to intestinal lymphoma.
For personal use. Only reproduce with permission from The Lancet Publishing Group.
Summary
Background Although previous studies have shown
increased mortality in patients with coeliac disease and their
relatives, no data are available in relation to different
patterns of clinical presentation. We assessed mortality in
patients with coeliac disease and their first-degree relatives.
Methods We enrolled, in a prospective cohort study, 1072
adult patients with coeliac disease consecutively diagnosed
in 11 gastroenterology units between 1962 and 1994, and
their 3384 first-degree relatives. We compared the number of
deaths up to 1998 with expected deaths and expressed the
comparison as standardised mortality ratio (SMR) and
relative survival ratio.
Findings 53 coeliac patients died compared with 25·9
expected deaths (SMR 2·0 [95% CI 1·52·7]). A significant
excess of mortality was evident during the first 3 years after
diagnosis of coeliac disease and in patients who presented
with malabsorption symptoms (2·5 [1·83·4]), but not in
those diagnosed because of minor symptoms (1·1 [0·52·2])
or because of antibody screening (1·2 [0·17·0]). SMR
increased with increasing delay in diagnosis and for patients
with poor compliance with gluten-free diet. Non-Hodgkin
lymphoma was the main cause of death. No excess of
deaths was recorded in relatives with coeliac disease.
Interpretation. Prompt and strict dietary treatment
decreases mortality in coeliac patients. Prospective studies
are needed to clarify the progression of mild or symptomless
coeliac disease and its relation to intestinal lymphoma.
Lancet 2001; 358: 35661
Introduction
Findings from population-based studies have shown that
the true frequency of coeliac disease is high
1
even in
countries where it was thought to be rare.
2
An increased
overall and cancer mortality has been reported in adult
patients with coeliac disease
3,4
and their relatives,
5
which
lends support to the clinical importance of this disorder.
Improved knowledge of the wide clinical spectrum of
coeliac disease
6,7
and the use of powerful screening tests
8,9
have radically changed the pattern of presentation of this
disorder. Previous mortality could be underestimated by
the inclusion of subclinical and symptom-free patients,
but no information is available on the prognosis of the
new forms of coeliac disease. Additionally, the reported
1·9-fold
3
and 3·4-fold
4
increases in mortality might be
excessive because some patients with a less favourable
outcome, such as those with refractory sprue and
intestinal lymphoma, were probably enrolled in those
series.
10
The finding that the mortality excess is mainly
accounted for by deaths occurring within a short time
after diagnosis
3
indirectly lends support to this possibility.
We did a prospective study to find out whether
differences in patterns of clinical presentation of coeliac
disease are associated with differences in prognosis;
whether age at diagnosis, time between onset of symptoms
that subsequently led to intestinal biopsy (diagnostic
delay), and degree of adherance with a gluten-free diet
affect mortality; and whether first-degree relatives of these
patients have an increased risk of mortality.
Methods
Patients
We identified patients with biopsy-proven coeliac disease
from the records of the 11 gastroenterology units. The
diagnostic criteria were subtotal or severe partial villous
atrophy and crypt hyperplasia, and histological
improvement after gluten-free diet. We enrolled 1072
consecutive patients older than 18 years at the time of
diagnosis, between Jan 1, 1962, and 31 Dec, 1994, at
11 gastroenterology units. These units, evenly distributed
through Italy, were selected from those participating in
the Italian Club del Tenue study in accordance with the
following inclusion criteria: completeness of clinical
records (all the diagnoses of coeliac disease reported from
the start of specific diagnostic activity), and reliability of
the diagnoses throughout the entire period of activity.
Units flagged patients by comparing records with the
corresponding small-bowel pathology lists.
We obtained information on vital status, sex, age at
time of diagnosis, date of initial presentation, diagnostic
delay (time from onset of symptoms to intestinal biopsy),
and dietary adherence (adherent or not adherent)
recorded at presentation or during subsequent clinical
surveillance. We classified patients into three subtypes of
coeliac disease according to clinical presentation:
11
severe,
with symptoms of malabsorption such as diarrhoea,
weight loss, or both that led the patient to seek medical
care; mild, with only trivial, transient, or seemingly
ARTICLES
356
THE LANCET • Vol 358 • August 4, 2001
Mortality in patients with coeliac disease and their relatives:
a cohort study
Giovanni Corrao, Gino Roberto Corazza, Vincenzo Bagnardi, Giovanna Brusco, Carolina Ciacci, Mario Cottone,
Carla Sategna Guidetti, Paolo Usai, Pietro Cesari, Maria Antonietta Pelli, Silvano Loperfido, Umberto Volta,
Antonino Calabró, Maria Certo, for the Club del Tenue Study Group
Cattedra di Statistica Medica, Università di Milano-Bicocca,
20126 Milano, Italy (Prof G Corrao
PhD, V Bagnardi PhD); Cattedra di
Gastroenterologia, IRCCS Policlinico San Matteo, Università di
Pavia (Prof G R Corazza
MD, G Brusco MD); Cattedra di
Gastroenterologia, Università ‘Federico II’ di Napoli (C Ciacci
MD);
Clinica Medica ‘R’, Università di Palermo (M Cottone
MD); Cattedra
di Gastroenterologia, Università di Torino
(Prof C Sategna Guidetti
MD); Clinica Medica, Università di Cagliari
(P Usai
MD); Unita di Gastroenterologia, Casa di Cura Poliambulanza
di Brescia (P Cesari
MD); Cattedra di Gastroenterologia, Università
di Perugia (M A Pelli
MD); Divisione di Gastroenterologia, Ospedale
di Treviso (S Loperfido
MD); Cattedra di Medicina Interna, Università
di Bologna (U Volta
MD); Cattedra di Gastroenterologia, Università
di Firenze (A Calabró
MD); and Cattedra di Medicina Interna,
Università Cattolica di Roma (M Certo
MD)
Correspondence to: Prof Giovanni Corrao
(e-mail: giovanni.corrao@unimib.it)
For personal use. Only reproduce with permission from The Lancet Publishing Group.
unrelated symptoms; and symptomless, which included
symptom-free patients diagnosed during antibody
screening.
Data collection
We interviewed the patients in the first few months of
1999 by telephone and asked about adherence to their
gluten-free diet from the date of coeliac disease diagnosis.
We classified diet as: gluten-free (no gluten intake); low
gluten (occasional lapses, ie, 2–3 lapses per month); or
containing gluten (frequent lapses, ie, once or more per
week or no gluten restriction at all).
12
Further questions
were about the date of birth, coeliac disease diagnosis,
vital status, or date of death of first-degree relatives. For
patients who had died, we obtained information from
relatives or people who used to live with them. Date and
place of death of these cases were also recorded.
Researchers from eight of the 11 units agreed to interview
873 patients. Eight patients, or their relatives, could not
be traced and three did not respond to the interview.
Therefore 862 (80%) of 1072 patients were included in
the second part of our study.
Cohort follow-up
We separated patients and their relatives into three
cohorts. The first included all the 1072 patients with
coeliac disease (patient cohort), for whom follow-up
started at the time of diagnosis. The second included the
parents of 862 patients (parent cohort), for whom follow-
up started at the birth of the son or daughter with coeliac
disease. The third group included all the siblings of 862
patients (sibling cohort), for whom follow-up started at
birth. When two or more relatives were diagnosed in the
same gastroenterology unit, we included the patient
diagnosed first in the patient cohort, and the other patient
or patients in the parent or sibling cohort. The end of
follow-up was fixed at Dec 31, 1998, for all cohorts.
We ascertained the vital status of patients in all cohorts
from the last known municipality of residence when this
information could not be obtained from clinical records or
interview. Copies of the death certificate for all patients
who had died were obtained from the local mortality
registry, and the cause of death was classified and coded
according to the ninth revision of the International
Classification of Diseases. For two unavailable death
certificates, we obtained the causes of death from the
death file from the Italian National Institute of Statistics.
Data analysis
We calculated the expected number of deaths in every
unit by multiplying the number of patient-years in each
stratum of sex, 5-year age-groups, and calendar year by
the corresponding mortality rates. These rates were
reported for each region covered by the units starting from
1960 to 1964 (patient cohort) or for the entire Italian
population starting from 1900 to 1904 (parent and sibling
cohorts). We calculated the standardised mortality ratios
(SMRs [ratio of observed to expected deaths]), taking into
account all-cause mortality and selected causes of death
for the patient cohort, and all causes of death for the
parent and sibling cohorts. The corresponding 95% CIs
were calculated on the assumption that the observed
number of deaths had a Poisson distribution. We stratified
SMRs according to a series of demographic and clinical
classifications and assessed differences by testing their
heterogeneity between sex, pattern of presentation, and
adherence with gluten-free diet, or for trend over strata
(calendar year of initial presentation, age at diagnosis, and
diagnostic delay) according to Breslow and Day’s
analysis.
13
Power calculations had suggested that the study
would give an 80% chance of detecting significant overall
SMRs higher than 1·6 (patient cohort), 1·2 (parent
cohort), and 1·3 (sibling cohort), and a 2·3-fold higher
SMR in patients with severe coeliac disease than in those
with mild disease.
13
We compared observed and expected cumulative
overall survival proportions within each cohort. Expected
proportions were calculated according to the national sex,
age, and calendar-year life tables.
14
We presented results
as relative survival ratios, and the corresponding 95% CI
were based on the normal approximation of the binomial
distribution. We used several statistical tests (
2
and its
version for the trend, parametric or non-parametric one-
way analysis of variance) when appropriate to test
differences or trends in distribution of a factor between
the classifications of another factor and among
gastroenterology units. For all hypotheses tested, two-
tailed p values less than 0·05 were deemed significant.
Results
Two of the 11 gastroenterology units flagged patients
from before 1974, four flagged from 1975 to 1984, and
the remaining five flagged after 1985. Median number of
patients per unit was 48 (IQR 29–144). Complete data
were available except for diagnostic delay and dietary
adherence from 68 and 89 clinical records, respectively.
Distributions of patients’ sex, age at diagnosis, diagnostic
delay, pattern of presentation, and adherence with gluten-
free diet did not significantly differ between units. In the
first cohort that included all 1072 coeliac disease patients,
the ratio of men to women was 1/3; mean age at diagnosis
35·7 years (SD 14·1); mean follow-up 6·0 years (SD 4·9);
and median diagnostic delay 17 months (IQR 4–122).
Table 1 shows mortality for the patient cohort by
demographic classifications. 50 patients were lost during
follow-up. An increase in mortality was noted in the
entire patient cohort. The overall SMR did not differ by
sex, age at diagnosis, or year of presentation. By contrast,
SMRs increased significantly for patients with diagnostic
delay more than 1 year, and we noted a significant
association between increasing SMR and diagnostic
delay. Nearly 50% of patients were diagnosed with mild
or symptomless coeliac disease. There was a significant
excess of deaths among patients who presented with only
malabsorption symptoms. No excess mortality was seen
in patients with mild or symptomless coeliac disease, and
SMRs were heterogeneous between classifications of
clinical patterns.
Significant excess of mortality was seen in patients who
did not adhere to a gluten-free diet, recorded from clinical
records (SMR 10·7 [95% CI 6·0–17·1]) and interview
(6·1 [4·2–8·6]). However, the methods of assessing
adherence to a gluten-free diet showed an important
disparity. 95 (10%) of 915 patients who reported good
adherence in clinical records admitted frequent dietary
lapses at the direct interview. However, 16 (24%) of 68
patients who were classified as non-adherent in clinical
records turned out to be strictly adherent at interview. For
627 patients, interview and clinical records showed
probable adherence. For 155, the two methods showed
agreement for probable non-adherence. The status of
adherence was unknown for 290 patients because they
were not searched for, traced, or did not attend interview,
and were classified as uncertain. After this new
classification, SMRs differed between classifications, and
significant excesses of mortality were noted in probable
non-adherent patients and in patients classified as
uncertain.
ARTICLES
THE LANCET • Vol 358 • August 4, 2001
357
For personal use. Only reproduce with permission from The Lancet Publishing Group.
SMR was increased for patients with a diagnostic delay
of 120 months or more (2·5 [1·3–4·6), severe coeliac
disease (2·9 [1·8–4·3), and poor adherence to a gluten-
free diet (5·2 [3·4–7·8]) in the 887 patients who were
diagnosed as having coeliac disease after 1985. In the
same way, SMR was increased for patients with a
diagnostic delay of 120 months or more (3·3 [1·5–6–2])
and poor adherence to gluten-free diet (5·5 [2·5–10·5) in
the 590 patients with severe coeliac disease. This analysis
was not done for patients with mild coeliac disease,
because only nine deaths occurred in this group.
The number of patients decreased progressively from
the earliest to the latest period for diagnostic delay of 120
months or more (15 [52%], 39 [25%], and 219 [25%];
p=0·028), malabsorption symptoms (26 [90%], 126
[80%], and 438 [49%]; p=0·0007), and for likely non-
adherence (11 [38%], 40 [26%], and 104 [12%];
p=0·0005).
The number of patients who presented with a
diagnostic delay of 120 months or more did not differ
significantly between clinical pattern (153 [26%] and 20
[25%] in patients with severe and mild coeliac disease,
respectively; p=0·70), and adherence classification (163
[26%] likely, 42 [27%] non-adherent, and 68 [23%]
uncertain; p=0·63). Finally, the number of patients who
were probable non-adherents did not vary significantly
between clinical pattern (97 [16%] and 98 [12%] in those
with severe and mild disease, respectively; p=0·91).
Table 2 shows the causes of death recorded in the
patient cohort. There was a significant excess of deaths
from malignant disease and from diseases of the
respiratory and digestive tracts. Of the patients who died
as a result of malignant disease, non-Hodgkin lymphoma
was mentioned in two-thirds of death certificates; the
other eight deaths from malignant disease did not exceed
the number expected. The sites of these cancers were the
stomach in two patients, small intestine in one, liver in
two, pancreas in one, pleura in one, and unspecified
leukaemia in one. Of the deaths from respiratory
diseases, viral or bacterial pneumonia was mentioned on
three death certificates, lung emphysema on one, and
fibrosing alveolitis on one. Of the deaths due to diseases
of the digestive system, ulcerative jejunoileitis and coeliac
disease were mentioned on five and three certificates,
respectively. Other deaths due to digestive diseases
were not significantly in excess (SMR 1·7 [95% CI
0·6–3·7]).
Figure 1 shows the trend in relative survival ratios of the
entire cohort of coeliac disease patients. Reductions were
noted within the first 3 years after diagnosis (relative
ARTICLES
358
THE LANCET • Vol 358 • August 4, 2001
Patients Patient-years Observed deaths Expected deaths SMR (95% CI) p
Overall 1072 (100%) 6444 53 25·92·0 (1·52·7) <0·0001
Sex
Men 258 (24%) 1446 22 11·02·0 (1·33·0) 0·004
Women 814 (76%) 4998 31 14·92·1 (1·43·0) 0·0003
Test for heterogeneity: p=0·99
Age at diagnosis (years)
1829 373 (35%) 2080 3 1·22·5 (0·57·3) 0·24
3049 507 (47%) 3293 14 5·92·4 (1·34·0) 0·006
50 192 (18%) 1071 36 18·81·9 (1·32·6) 0·0005
Test for trend: p=0·46
Year of presentation
196274 29 (3%) 563 8 2·53·2 (1·46·3) 0·008
197584 156 (15%) 2002 14 7·61·8 (1·03·1) 0·05
1985 887 (83%) 3879 31 15·82·0 (1·32·8) 0·0009
Test for trend: p=0·38
Diagnostic delay (months)
12 344 (32%) 2081 19 12·71·5 (0·92·3) 0·12
12119 320 (30%) 2030 19 7·32·6 (1·64·1) 0·0004
120 273 (25%) 1706 15 3·93·8 (2·26·4) <0·0001
Unknown or not applicable* 135 (13%) 627 0 2·0
Test for trend: p=0·004
Pattern of presentation
Severe 590 (55%) 4255 43 17·22·5 (1·83·4) <0·0001
Mild 415 (39%) 1870 9 7·91·1 (0·52·2) 0·79
Symptomless 67 (6%) 319 1 0·81·2 (0·17·0) 0·99
Test for heterogeneity: p=0·033
Adherence to gluten-free diet
Likely 627 (59%) 3794 5 10·50·5 (0·21·1) 0·16
Not likely 155 (15%) 998 26 4·36·0 (4·08·8) <0·0001
Uncertain 290 (27%) 1652 22 11·12·0 (1·23·0) 0·005
Test for heterogeneity: p<0·0001
SMR=standardised mortality ratio. *Unknown in 68 patients (clinical records lacking this information) and not applicable in 67 patients with symptomless disease.
Test for trend does not include this category.
Table 1: Demographics, clinical features, and overall mortality of patient cohort
Causes of death (ICD9 codes) Observed deaths Expected deaths SMR (95% CI) p
Malignant diseases (140208) 24 9
.
12
.
6 (1
.
73
.
9) <0·0001
Non-Hodgkin lymphoma (200 or 202) 16 0
.
269
.
3 (40
.
7112
.
6) <0·0001
Other malignancies 8 8
.
90
.
9 (0
.
41
.
8) 0·94
Circulatory system diseases (390459) 7 9
.
90
.
7 (0
.
31
.
5) 0·46
Respiratory system diseases (460519) 5 1
.
43
.
6 (1
.
18
.
4) 0·03
Digestive system diseases (520579) 11 1
.
86
.
1 (3
.
010
.
9) <0·0001
Other causes of death 6 3
.
71
.
6 (0
.
63
.
5) 0·34
Table 2: Causes of death in patient cohort
For personal use. Only reproduce with permission from The Lancet Publishing Group.
survival ratio 0·98 [95% CI 0·97–0·98]). After this time,
the ratio between observed and expected long-term
survival probabilities stabilised (0·98 [0·96–0·99]). This
finding is consistent with a 2% higher cumulative
mortality in patients with coeliac disease than in the
Italian population, and implies that the excess of mortality
in these patients does not occur 3 years after diagnosis.
Figure 2 shows the comparison between trends in relative
survival ratios in patients classified according to the
pattern of clinical presentation. Significant reductions in
relative survival ratios were seen only for patients with
severe symptoms at diagnosis, with a trend similar to that
described for the entire cohort (0·97 [0·96–0·99] at
3 years from diagnosis; 0·97 [0·95–0·99] at 8 years). By
contrast, observed and expected survivals did not differ
significantly in the patients with mild coeliac disease. This
analysis was not done for symptom-free patients, since
only one death occurred in this group.
Table 3 shows the mortality of first-degree relatives of
coeliac disease patients. Of 3384 relatives, 749 died
during the follow-up period, and 741·6 deaths were
expected. SMR did not differ significantly between
relatives. 123 (4%) relatives were diagnosed with coeliac
disease during their life. However, for these relatives, no
significant difference between observed and expected
deaths was noted.
Discussion
Our study on the mortality of coeliac disease was large
with respect to the number of patients enrolled.
Although the total number of patient-years at risk
accumulated by our cohort (6444) was lower
than that noted by Logan and colleagues
3
(8823), this
difference is unlikely to have affected the accuracy of our
mortality estimates, because previous studies
3,15
and our
study agree that excess in mortality risk occurs mainly
within the first years from diagnosis. Additionally, we have
noted a relation between mortality and the pattern of
clinical presentation, delay in diagnosis, and the degree of
compliance with a gluten-free diet.
1072 adult patients with coeliac disease were
consecutively diagnosed by 11 Italian gastroenterology
units, in which researchers had similar diagnostic criteria
for coeliac disease. This similarity was confirmed by the
absence of significant differences for the variables shown in
table 1. Because this study is not population-based, our
design might have generated a selection bias. However, this
possibility seems unlikely because the observed distribution
of the clinical pattern of coeliac disease at presentation and
of demographic factors did not differ from those reported
by a large Italian multicentre survey.
16
Furthermore,
compared with the unique population-based investigation,
3
ARTICLES
THE LANCET • Vol 358 • August 4, 2001
359
1·02
1·00
0·98
0·96
0·94
Relative survival ratio
0123
Years from diagnosis
45 86 7
Figure 1: Relative survival ratios and corresponding 95% CI for
1072 patients with coeliac disease
Continuous line=relative survival ratio. Broken line=95% CI.
Figure 2: Relative survival ratios and corresponding 95% CI for
patients with severe (upper) and mild (lower) coeliac disease
Continuous line=relative survival ratio. Broken line=95% CI.
1·04
1·02
1·00
0·98
0·96
0·94
Relative survival ratio
0123
Years from diagnosis
45 86 7
1·02
1·00
0·98
0·96
0·94
Number Person-years Observed deaths Expected deaths SMR (95% CI) p
Fathers
All 862 26 673 337 340
.
71
.
0 (0
.
91
.
1) 0·87
With ascertained coeliac disease 22 659 7 8
.
40
.
8 (0
.
31
.
7) 0·80
Mothers
All 862 29 538 259 243
.
31
.
1 (0
.
91
.
2) 0·33
With ascertained coeliac disease 15 466 4 3
.
71
.
1 (0
.
32
.
7) 0·82
Brothers
All 822 27 655 80 72
.
31
.
1 (0
.
91
.
4) 0·39
With ascertained coeliac disease 27 1061 3 2
.
71
.
1 (0
.
23
.
3) 0·79
Sisters
All 838 35 153 73 85
.
30
.
9 (0
.
71
.
1) 0·20
With ascertained coeliac disease 59 2327 5 5
.
70
.
9 (0
.
32
.
1) 0·82
SMR=standardised mortality ratio.
Table 3: Mortality in 3324 first-degree relatives of 862 patients with coeliac disease
For personal use. Only reproduce with permission from The Lancet Publishing Group.
we reported a lower mortality rate (8·2 vs 13·0 per 1000
patients per year) but an almost identical SMR (2·0 vs 1·9).
We therefore confirm a two-fold increase in overall
mortality of adult patients with coeliac disease
3
and that this
increase accounts for deaths within a short time from
diagnosis.
3,15
These findings were not due to inclusion of
patients who had disorders other than coeliac disease,
because regrowth of villi after gluten-free diet was initially
documented in all patients except for those who admitted a
poor compliance with treatment from the beginning of the
study. Our results were consistent with the fact that the
duration of coeliac disease is often short before the onset or
diagnosis of lymphoma.
17
Because most patients who died
within 3 years of coeliac disease diagnosis had non-
Hodgkin lymphoma, ulcerative jejunitis (regarded as a
malignant disease strictly related to lymphoma)
18
or other
tumours, the recorded excess of mortality during this time
could be due to the occurrence of malignant disease, rather
than to coeliac disease alone. After this time, the patients
with coeliac disease who survived showed a probability of
death similar to that of the general Italian population.
Although the severity of clinical presentation should
relate to morbidity rather than to mortality of coeliac
disease, most of the increased mortality (43 of 53 deaths)
occurred in patients who presented with classic symptoms
of malabsorption at diagnosis. By contrast, patients
diagnosed with mild symptoms or by antibody screening
did not show any relevant excess of mortality. These results
are lent support by findings from two recent studies
showing that patients with dermatitis herpetiformis (a
disorder characterised by bullous skin disease, gluten-
sensitive villous atrophy, and minor or absent
malabsorption symptoms) have no increased general
mortality.
19,20
Since symptom-free patients with coeliac
disease are undetected and untreated for a long period of
time, they might have an increased likelihood of developing
lymphomas.
21
This possibility does not necessarily
contradict our results. Patients who were originally
subclinical or symptom free might have subsequently
developed malabsorption symptoms (therefore classified as
having a severe form of coeliac disease) because they
already had a lymphoma (or other malignant disease) at the
time of diagnosis, even though they transiently maintained
their capacity to respond to gluten-free diet.
22
We noted a positive association between diagnostic delay
and SMR that rises to 3·8 in patients diagnosed
10 years or more after the onset of symptoms. Our results
emphasise the need for prompt diagnosis and treatment
also in those patients with a minor or symptomless form of
coeliac disease. Nielsen and colleagues
4
did not show any
difference in mortality risk for adult coeliac disease patients
according to their adherance to gluten-free diet. However,
Logan and colleagues,
3
noted a non-increased mortality
rate in adult patients who had been on a strict gluten-free
diet since diagnosis at childhood. Collin and colleagues,
23
reported a normal survival in a cohort of adult patients,
83% of whom adhered strictly to a gluten-free diet. Our
results lend support to these studies because we paid
greater attention to the reliability of the information on diet
adherence.
We suspect that clinicians arbitrarily report poor
adherence for patients who drop out because of death,
whereas they probably report good adherence for those
under closer medical surveillance. This suggestion could
explain the high SMR (10·7) in patients classified as non-
adherent on the basis of clinical records. However, the
reliability of the information obtained directly from the
patients is questionable
24
and the same might apply to
interviews of relatives or collaterals of patients not traceable
or who died during follow-up. Because of this unreliability,
we expressed the degree of adherence to gluten-free diet by
comparing the available information sources. All other
dietary states were judged as uncertain or unknown.
By contrast with a previous study,
5
relatives of patients
with coeliac disease did not show any excess of mortality
compared with the general Italian population. Assuming
that 338 (10%) of the 3384 relatives were affected by
coeliac disease
25
and that this proportion of relatives had a
two-fold increase in mortality, we should have seen 812
deaths. Of the 338 assumed coeliac patients, coeliac disease
was diagnosed in only 123. Surprisingly, the SMR for these
relatives was close to 1. However, mortality occurrence in
relatives referred to the period after the birth of the index
case (parents) or to their entire life (brothers and sisters).
Thus, putative excesses of mortality occurring within a
short time after diagnosis, were not appreciable for such a
long follow-up period.
Our results confirm that the overall mortality in adult
coeliac disease patients is double that of the general
population, and show that delay in diagnosis, poor
adherence to treatment, and severity of symptoms at
presentation unfavourably affect patients’ outlook.
Prospective studies are needed to elucidate the progression
of mild or symptomless coeliac disease and their relation to
intestinal lymphoma. However, the normal survival of
patients with these forms of disease should not discourage
either an active search for these patients or their strict and
life-long avoidance of dietary gluten.
The patients who died within a short time of diagnosis
might have originally had a subclinical or symptomless form
of the disease. Coeliac disease might have been
subsequently unmasked by the onset and growth of
malignant disease, and therefore the patient could have
benefited from the protective effect of a gluten-free diet at
an earlier stage.
26
Gluten-free diet, as well as decreasing
mortality risk, has recently been shown to improve the
quality of life even in patients with symptomless coeliac
disease.
27
Contributors
G Corrao and G R Corazza co-ordinated the study, designed the protocol,
supervised the data at the co-ordinating centre, and wrote and edited the
paper. V Bagnardi dealt with the mortality follow-up and collaborated with
G Corrao in the statistical analyses. G Brusco, C Ciacci, M Cottone, C
Sategna Guidetti, P Usai, P Cesari, M A Pelli, S Loperfido, U Volta,
A Calabró, and M Certo obtained the data at individual study sites.
Acknowledgments
We thank the Associazione Italiana Celiachia, the Istituto Superiore di
Sanità (research project “Prevention of risk factors of maternal and child
health”) and the Ministero dell’Università e della Ricerca Scientifica e
Tecnologica (research project “Autoimmunity and coeliac disease: gluten,
induction of autoimmune phenomena and effector mechanisms of intestinal
damage”) for financial support.
References
1 Catassi C, Rätsch I-M, Fabiani E, et al. Coeliac disease in the year
2000: exploring the iceberg. Lancet 1994; 343: 200–03.
2 Berti I, Horvath K, Green PH. Prevalence of celiac disease among risk
groups and the general population in USA. Gastroenterology 2000; 118:
A696 (abstr).
3 Logan RFA, Rifkind EA, Turner ID, Ferguson A. Mortality in coeliac
disease. Gastroenterology 1989; 97: 265–71.
4 Nielsen OH, Jacobsen O, Pedersen ER, et al. Non-tropical sprue:
malignant disease and mortality rate. Scand J Gastroenterol 1985; 20:
13–18.
5 Stokes PL, Prior P, Sorahan TM, McWalter RJ, Waterhouse JAH,
Cooke WT. Malignancy in relatives of patients with coeliac disease.
Br J Prev Soc Med 1986; 30: 17–21.
6 Logan RFA, Tucker G, Rifkind EA, Heading RC, Ferguson A. Changes
in clinical features of coeliac disease in adults in Edinburgh and the
Lothians 1960–79. BMJ 1983; 286: 95–97.
ARTICLES
360
THE LANCET • Vol 358 • August 4, 2001
For personal use. Only reproduce with permission from The Lancet Publishing Group.
7 Collin P, Reunala T, Rasmussen M, et al. High incidence and
prevalence of adult coeliac disease: augmented diagnostic approach.
Scand J Gastroenterol 1997, 32: 1129–33.
8 Hin H, Bird G, Fischer P, Mahy N, Jewell D. Coeliac disease in primary
care: case finding study. BMJ 1999; 318: 164-67.
9 Corazza GR, Valentini RA, Frisoni M, et al. Gliadin immune reactivity
is associated with overt and latent enteropathy in relatives of coeliac
patients. Gastroenterology 1992; 103: 1517–22.
10 Trier JS. Celiac sprue. N Engl J Med 1991; 325: 1709–19.
11 Swinson CM, Levi AJ. Is coeliac disease underdiagnosed? BMJ 1980;
281: 1258–60.
12 Greco L, Mayer M, Ciccarelli G, Troncone R, Auricchio S. Compliance
to a gluten-free diet in adolescents or what do 300 coeliac adolescents
eat every day? Ital J Gastroenterol Hepatol 1997; 29: 305–11.
13 Breslow NE, Day NE. Statistical methods in cancer research, vol II: the
design and analysis of cohort studies. Lyon: International Agency for
Research on Cancer, 1987.
14 Thomsen BL, Keiding N, Altman DG. A note on the calculation of
expected survival, illustrated by the survival of liver transplant patients.
Stat Med 1991; 10: 733–38.
15 Cottone M, Termini A, Oliva L, et al. Mortality and causes of death in
celiac disease in a Mediterranean area. Dig Dis Sci 1999; 44: 2538–41.
16 Bottaro G, Cataldo F, Rotolo N, Spina M, Corazza GR. The clinical
pattern of subclinical/silent celiac disease: an analysis of 1026
consecutive cases. Am J Gastroenterol 1999; 94: 691–96.
17 Wright DH. The major complications of coeliac disease. Baillière’s Clin
Gastroenterol 1995; 9: 351–69.
18 Ashton-Key M, Diss TC, Pan L, Du MQ, Isaacson PG. Molecular
analysis of T-cell clonality in ulcerative jejunitis and
enteropathy-associated T-cell lymphoma. Am J Pathol 1997; 151:
493–98.
19 Swerdlow AJ, Whittaker S, Carpenter IM, English JSC. Mortality and
cancer incidence in patients with dermatitis herpetiformis: a cohort
study. Br J Dermatol 1993; 129: 140–44.
20 Collin P, Pukkala E, Reunala T. Malignancy and survival in dermatitis
herpetiformis: a comparison with coeliac disease. Gut 1996; 38: 528–30.
21 Howell WM, Leung ST, Jones DB, et al. HLA-DRB, -DQA, and
-DQB polymorphism in celiac disease and enteropathy associated
T-cell lymphoma: common features and additional risk factors for
malignancy. Hum Immunol 1995; 43: 29–37.
22 Freeman HJ, Weinstein WM, Shnitka TK, Piercey JRA, Wensel RH.
Primary abdominal lymphoma: presenting manifestation of celiac sprue
or complicating dermatitis herpetiformis. Am J Med 1977; 63: 585–94.
23 Collin P, Reunala T, Pukkala E, Laippala P, Keyriläinen O, Pasternack
A. Coeliac disease: associated disorders and survival. Gut 1994; 35:
1215–18.
24 Baker PG, Barry RE, Read AE. Detection of continuing gluten ingestion
in treated coeliac patients. BMJ 1975; 1: 486–88.
25 Auricchio S, Mazzacca G, Tosi R, Visakorpi J, Mäki M, Polanco I.
Coeliac disease as a familial condition: identification of asymptomatic
coeliac patients within family groups. Gastroenterol Int 1988; 1:
25–31.
26 Holmes GKT, Prior P, Lane MR, Pope D, Allan RN. Malignancy in
coeliac disease—effect of a gluten free diet. Gut 1989; 30:
333–38.
27 Mustalahti K, Lohiniemi S, Laippala P, Collin P, Mäki M.
Improvement of the quality of life of silent celiac disease patients during
the gluten-free diet warrants screening. Gastroenterology 2000; 118:
A369.
ARTICLES
THE LANCET • Vol 358 • August 4, 2001
361
    • "among patients with poor adherence to a GFD [26] . West and others found that the overall HR for mortality in CD was 2.09 in the first year after diagnosis and declined to 1.1 after the first year of diagnosis [27] . Meta-analysis done by Tio et al [28] showed an increase risk for all-cause mortality in CD patients with an OR of 1.24 (95%CI: 1.19-1.3). "
    Article · Jan 2016
    • "In particular, we confirm that the complications of CD are burdened by a high rate of mortality and that death occurs soon after the diagnosis of the complication (Figure 2). We also confirmed that coeliac patients who do not adhere to a strict GFD, with symptoms of classic CD at the onset and with a diagnosis late in life are at greater risk of developing these complications [18]. "
    [Show abstract] [Hide abstract] ABSTRACT: Background Coeliac disease is a common enteropathy characterized by an increased mortality mainly due to its complications. The natural history of complicated coeliac disease is characterised by two different types of course: patients with a new diagnosis of coeliac disease that do not improve despite a strict gluten-free diet (type A cases) and previously diagnosed coeliac patients that initially improved on a gluten-free diet but then relapsed despite a strict diet (type B cases). Our aim was to study the prognosis and survival of A and B cases. Methods Clinical and laboratory data from coeliac patients who later developed complications (A and B cases) and sex- and age-matched coeliac patients who normally responded to a gluten-free diet (controls) were collected among 11 Italian centres. Results 87 cases and 136 controls were enrolled. Complications tended to occur rapidly after the diagnosis of coeliac disease and cumulative survival dropped in the first months after diagnosis of complicated coeliac disease. Thirty-seven cases died (30/59 in group A, 7/28 in group B). Type B cases presented an increased survival rate compared to A cases. Conclusions Complicated coeliac disease is an extremely serious condition with a high mortality and a short survival. Survival depends on the type of natural history.
    Full-text · Article · Aug 2014
    • "CD affects women more frequently than men; one study found a female-to-male ratio of 3:1 in adult CD [5]. CD is associated with increased morbidity and mortality , partially due to higher risk of lymphomas678910. A gluten-free diet is the standard treatment, and strict adherence usually results in clinical, serological, and histological remission. "
    [Show abstract] [Hide abstract] ABSTRACT: Patients with celiac disease (CD) have low bone mineral density. Evidence of increased fracture risk in these patients is conflicting, and the indication for bone mineral density screening of all adult CD patients is debated. Our aim was to review current published data on fractures in CD. Cross-sectional cohort studies and one case study were identified by searching Medline and Embase. Although the identified studies are heterogeneous and difficult to compare, the overall findings indicate a positive association between CD and risk of fracture. Adult patients with CD should be considered for bone densitometry in order to estimate fracture risk.
    Full-text · Article · Apr 2014
Show more