Patient Perception of Treatment Burden Is High in Celiac Disease Compared With Other Common Conditions

Abstract

Objectives:
The only treatment for celiac disease (CD) is life-long adherence to a gluten-free diet (GFD). Noncompliance is associated with signs and symptoms of CD, yet long-term adherence rates are poor. It is not known how the burden of the GFD compares with other medical treatments, and there are limited data on the socioeconomic factors influencing treatment adherence. In this study, we compared treatment burden and health state in CD compared with other chronic illnesses and evaluated the relationship between treatment burden and adherence.

Methods:
Survey was mailed to participants with CD, gastroesophageal reflux disease (GERD), irritable bowel syndrome, inflammatory bowel disease, hypertension (HTN), diabetes mellitus (DM), congestive heart failure, and end-stage renal disease (ESRD) on dialysis. Surveys included demographic information and visual analog scales measuring treatment burden, importance of treatment, disease-specific health status, and overall health status.

Results:
We collected surveys from 341 celiac and 368 non-celiac participants. Celiac participants reported high treatment burden, greater than participants with GERD or HTN and comparable to ESRD. Conversely, patients with CD reported the highest health state of all groups. Factors associated with high treatment burden in CD included poor adherence, concern regarding food cost, eating outside the home, higher income, lack of college education, and time limitations in preparing food. Poor adherence in CD was associated with increased symptoms, income, and low perceived importance of treatment.

Conclusions:
Participants with CD have high treatment burden but also excellent overall health status in comparison with other chronic medical conditions. The significant burden of dietary therapy for CD argues for the need for safe adjuvant treatment, as well as interventions designed to lower the perceived burden of the GFD.

Full text

THE RED SECTION
© 2014 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
nature publishing group 1
Celiac disease (CD) is a chronic immune-mediated enteropathy
triggered by gluten-containing foods.  e prevalence is estimated
to be 1:70 – 1:300, and the only treatment is life-long adherence to
a gluten-free diet (GFD) ( 1 – 5 ). Prior studies estimate adherence
to the GFD to be as low as 36 – 45 % ( 6 – 8 ). CD di ers from many
other chronic diseases because dietary changes are the only current
therapy for disease management. Poor dietary compliance is the
leading cause of ongoing symptoms in participants with CD ( 9 ).
 is is especially concerning considering the numerous poten-
tial complications of untreated CD, including reduction in bone
mineral density, malignancy, and increased mortality rate ( 3,10,11 ).
Adherence to GFD has been associated with improvements in
quality of life( 12 – 15 ), bone mineral density ( 16 – 21 ), fatigue ( 22 ),
infertility ( 23 – 26 ), adverse pregnancy outcomes ( 27 – 29 ), and risk
of lymphoproliferative malignancy ( 30,31 ).
 ere are multiple potential factors accounting for low dietary
adherence in CD, including limited availability and higher cost
of gluten-free foods ( 32 – 34 ), reduced enjoyment of food ( 35 ),
and social isolation when dining out ( 35,36 ). Although there are
many unique features of the GFD that may reduce adherence, it
Patient Perception of Treatment Burden Is High
in Celiac Disease Compared With Other Common
Conditions
S v e t a S h a h , M D 1 , 2 , M o n a A k b a r i , M D , M P H 1 , 2 , Rohini Vanga , MD 1 , C i a r a n P . K e l l y , M D 1 , J o s h u a H a n s e n , M S 1 ,  i m m a i a h  eethira , MD 1 ,
S o h a i b Ta r i q , M D 1 , Melinda Dennis , MS, RD, LDN 1 a n d D a n i e l A . L e  e r , M D , M S 1
OBJECTIVES: The only treatment for celiac disease (CD) is life-long adherence to a gluten-free diet (GFD). Non-
compliance is associated with signs and symptoms of CD, yet long-term adherence rates are poor.
It is not known how the burden of the GFD compares with other medical treatments, and there are
limited data on the socioeconomic factors infl uencing treatment adherence. In this study, we com-
pared treatment burden and health state in CD compared with other chronic illnesses and evaluated
the relationship between treatment burden and adherence.
METHODS: Survey was mailed to participants with CD, gastroesophageal refl ux disease (GERD), irritable bowel
syndrome, infl ammatory bowel disease, hypertension (HTN), diabetes mellitus (DM), congestive heart
failure, and end-stage renal disease (ESRD) on dialysis. Surveys included demographic information
and visual analog scales measuring treatment burden, importance of treatment, disease-specifi c
health status, and overall health status .
RESULTS: We collected surveys from 341 celiac and 368 non-celiac participants. Celiac participants reported
high treatment burden, greater than participants with GERD or HTN and comparable to ESRD. Con-
versely, patients with CD reported the highest health state of all groups. Factors associated with high
treatment burden in CD included poor adherence, concern regarding food cost, eating outside the
home, higher income, lack of college education, and time limitations in preparing food. Poor adherence
in CD was associated with increased symptoms, income, and low perceived importance of treatment.
CONCLUSIONS: Participants with CD have high treatment burden but also excellent overall health status in compari-
son with other chronic medical conditions. The signifi cant burden of dietary therapy for CD argues
for the need for safe adjuvant treatment, as well as interventions designed to lower the perceived
burden of the GFD.
SUPPLEMENTARY MATERIAL is linked to the online version of the paper at http://www.nature.com/ajg
Am J Gastroenterol advance online publication, 1 July 2014; doi: 10.1038/ajg.2014.29
1 Celiac Center and Department of Gastroenterology, Beth Israel Deaconess Medical Center , Boston , Massachusetts , USA ;
2 The fi rst two authors contributed
equally to this work and are co-fi rst authors . Correspondence: Daniel A. Leffl er, MD, MS , Celiac Center and Department of Gastroenterology, Beth Israel
Deaconess Medical Center , 330 Brookline Ave. , Boston , Massachusetts 02215 , USA . E-mail: dleffl er@caregroup.harvard.edu

THE RED SECTION
The American Journal of GASTROENTEROLOGY VOLUME 104 | XXX 2014 www.amjgastro.com
2
is generally accepted that adherence is better with well-circum-
scribed treatments, such as medication administration, than
with health behaviors including dietary advice ( 37 ). Currently,
treatment of CD is limited exclusively to lifestyle modi cation,
which may contribute to poor adherence and elevated treat-
ment burden in comparison with many other chronic illnesses;
however, this hypothesis has not been rigorously evaluated.
Close follow-up with a physician, dietician, and support group
are routinely recommended in order to monitor and improve
dietary adherence ( 38,39 ), although the evidence for these inter-
ventions is limited.
To improve the quality of life and treatment adherence in CD,
a robust understanding of the burden of the GFD and the factors
that in uence adherence is necessary. However, relevant data
on this topic are limited ( 40 ) and in most studies have lacked
a validated measure of adherence.  e objective of this study
was to compare the treatment burden in CD with that of other
chronic illnesses and identify factors associated with treatment
burden and GFD adherence as measured by a validated survey
tool ( 41 ).
METHODS
Study subjects .  e study population consisted of participants
evaluated at the Celiac Center at Beth Israel Deaconess Medical
Center or who visited primary care clinics associated with Beth
Israel Deaconess Medical Center in Boston, MA.  is study was
approved by the Beth Israel Deaconess Medical Center Commit-
tee on Clinical Investigations.
Survey development and measures .  e survey was designed
to assess socioeconomic and demographic information, meas-
ures evaluating treatment burden and adherence, and speci c
questions relating health and treatment barriers. Visual analog
scales (VAS) have been used to study burden of disease across
a multitude of disorders, including gastroesophageal re ux
disease (GERD), Crohn ’ s disease, irritable bowel syndrome
(IBS), diabetes, and congestive heart failure (CHF) ( 42 – 46 ). We
adapted these existing surveys for the current study. Partici-
pants were asked to rate four domains on a scale of 0 – 100: (i)
di culty in following treatment; (ii) perceived importance of
following treatment; (iii) disease-speci c health; and (iv) overall
health (see Supplementary Appendix 1 online).  e question-
naire was  rst assessed for comprehensiveness and clarity by
patients with and without CD before being sent to participants.
For di culty in following treatment, a score of 0 indicated that
treatment is “ very easy ” and a score of 100 indicated that treat-
ment is “ very di cult ” . For importance of following treatment,
0 indicated “ not important ” and 100 indicated “ very important. ”
For disease-speci c health and overall health, 0 indicated “ worst
imaginable ” and 100 indicated “ best imaginable ” health states.
In addition to the VAS scales, participants with CD also com-
pleted the Celiac Dietary Adherence Test, a validated measure
of adherence to the GFD ( 41 ), and the Celiac Symptoms Index,
to measure CD symptoms ( 47 ).
Celiac cohort . To be eligible for participation in the celiac cohort,
individuals had to be > 18 years of age with biopsy-con rmed CD
for more than 3 months, have a valid US home address, and have
cognitive ability and English pro ciency suitable for independent
completion of the surveys.
Non-celiac cohorts . To be eligible for participation, individuals
had to have cognitive ability and English pro ciency suitable
for independent completion of the surveys, be > 18 years of age,
and have a valid US address. Participants were chosen on the
basis of one of seven chronic illnesses including hypertension
(HTN), diabetes mellitus (DM), CHF, end-stage renal disease
(ESRD) requiring dialysis, GERD, IBS, and in ammatory bowel
disease (IBD). Diagnoses were preliminarily identi ed through
International Classi cation of Diseases, 9th edition codes and
con rmed in all cases through independent review of the medi-
cal record.  e presence of multiple illnesses did not preclude
inclusion. In these scenarios, participants were included only in
the illness diagnosis  rst used for identi cation.
Statistical analysis . Univariate statistics were used to evalu-
ate cohort means and s.d. Missing data were handled by cohort
response means. Student ’ s t -statistic and one-way analysis of
variance compared means between two groups and across three
or more groups, respectively. Post hoc sche e multiple compari-
son test was used to evaluate the di erences between each pair
of means. Stepwise linear regression analysis controlling for age
and gender was used to determine predictors of treatment bur-
den, GFD adherence, and perceived importance of treatment for
the celiac cohort. A P value of < 0.05 was considered statistically
signi cant. All analyses were performed using Stata (StataCorp
LP; College Station, TX).
RESULTS
Characteristics of study participants . Of 773 surveys mailed
to CD participants, 341 (45 % ) responded. Mean age at diagnosis
was 42.98 years (95 % con dence intervals (CI): 41.35, 44.61) and
mean age at time of the survey was 51.14 years (95 % CI: 49.53,
52.75). Participants had followed a GFD for a mean of 85.49
months (95 % CI: 77.63, 93.46). Of the 1,288 surveys mailed out to
non-CD participants, 368 (29 % ) responded. Baseline characteris-
tics are listed in Table 1 .
Celiac responders were younger than those with HTN ( P < 0.001),
GERD ( P = 0.001), DM ( P < 0.001), and CHF ( P < 0.001) and older
than those with IBD ( P = 0.008).  ere was no di erence in age
between the celiac cohort and those with ESRD ( P = 0.112) or IBS
( P = 0.903). Celiac responders were more likely to be women com-
pared with all other cohorts (HTN, P < 0.001; GERD, P < 0.001;
ESRD, P = 0.006; DM, P < 0.001; CHF, P = 0.003; IBD, P < 0.001;
IBS, P = 0.004) and were more likely to identify as Caucasian com-
pared with those with HTN ( P < 0.001), GERD ( P < 0.001), ESRD
( P < 0.001), DM ( P < 0.001), CHF ( P < 0.001), but not with IBD
( P = 0.052) or IBS ( P = 0.092). Celiac responders were more likely
to report education of college or greater compared with those

THE RED SECTION
© 2014 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
3
with HTN ( P < 0.001), GERD ( P = 0.015), ESRD ( P = 0.001), DM
( P < 0.001), CHF ( P < 0.001), but not with IBD ( P = 0.817) or IBS
( P = 0.647). Celiac responders were less likely to report income less
than $ 75,000 compared with those with HTN ( P < 0.001), ESRD
( P = 0.001), DM ( P < 0.001), or CHF ( P < 0.001), but not compared
with GERD ( P = 0.055), IBD ( P = 0.455), or IBS ( P = 0.579).
Celiac responders were less likely to visit specialists compared
with all other diagnoses (GERD, P = 0.006; ESRD, P < 0.001; DM,
P < 0.001; CHF, P < 0.001; IBD, P < 0.001; and IBS, P < 0.001),
except for those with HTN ( P = 0.105). Celiac responders were also
less likely to visit their primary care physician compared with par-
ticipants with ESRD ( P = 0.002), DM ( P < 0.001), CHF ( P < 0.001),
GERD ( P = 0.050), and IBS ( P -value = 0.011), but not when com-
pared with HTN ( P = 0.055) and IBD ( P = 0.948).
Treatment burden .  e mean reported burden of following a
GFD was 44.90 points on the VAS (0: very easy and 100: very dif-
 cult). In univariate regression analysis, patient factors that were
signi cantly associated with increased treatment burden were
poor adherence to GFD ( P < 0.001), increased severity of current
celiac symptoms ( P = 0.001), time limitations for the research,
purchase and preparation of foods ( P = 0.001), di culty with eat-
ing outside the home ( P = 0.005), concern with the cost of food
( P = 0.001), lack of college education ( P = 0.004), and hospitaliza-
tions within the past year ( P = 0.006). In the multivariate linear
regression model controlling for age and gender, time limitations
for the research, purchase and preparation of foods ( P = 0.010),
di culty with eating outside the home ( P = 0.023), concern with
food cost ( P = 0.012), lack of college education ( P = 0.010), and
poor adherence ( P < 0.001) remained signi cantly associated with
increased treatment burden. Income > $ 200,000 ( P = 0.017) was
also signi cant in multivariate analysis, but increased severity of
current celiac symptoms and hospitalizations were not ( Table 2 ).
Treatment burden, perceived treatment importance, disease-
speci c health, and overall health were compared between all
medical conditions assessed. Figures 1 – 4 graphically depict these
results.  e reported treatment burden for CD of 44.90 was higher
than all other groups in aggregate (mean: 33.01, P < 0.001). CD
also had the highest reported treatment burden of all conditions
assessed with the exception of ESRD, although this only reached
statistical signi cance for HTN (mean: 23.50, P < 0.001) and
GERD (mean score: 21.34, P < 0.001). Treatment burden was
highest overall for ESRD (mean: 56.41), followed by CD, DM
(mean: 41.74), IBS (mean: 40.38), CHF (mean: 38.46), and IBD
(mean: 31.91).
Adherence to GFD .  e mean Celiac Dietary Adherence Test
score in the CD cohort was 11.93 (95 % CI: 11.55, 12.31). In uni-
variate linear regression, poor adherence was associated with
income < $ 200,000 ( P = 0.047), unemployed status ( P = 0.050),
increased severity of current celiac symptoms ( P < 0.001), lower
perceived importance of treatment ( P < 0.001), and greater
treatment burden ( P < 0.001). In the multivariate model,
income < $ 200,000 ( P = 0.045), increased severity of current celiac
symptoms ( P < 0.001), and lower perceived importance of treat-
ment ( P < 0.001) remained associated with poor adherence a er
controlling for age and gender ( Table 3 ).
Perceived importance of following the GFD . M o s t p a r t i c i p a n t s
reported high importance of following a GFD, with a mean score
for treatment importance of 93.80 (95 % CI: 91.83 – 95.77). In uni-
variate linear regression, di culty with eating outside the home
( P = 0.026), female gender (94.89 vs. 90.29, P = 0.050), greater
adherence ( P < 0.001), and increased severity of current celiac
symptoms ( P = 0.024) were associated with higher perceived
importance of following a GFD. Only greater adherence ( P < 0.001)
and increased severity of current celiac symptoms ( P < 0.001) con-
tinued to have signi cance in the multivariate linear regression
model ( Table 4 ). In comparison with other diseases, only ESRD
rated treatment as more important (mean score: 94.67); however,
the di erence was only statistically signi cant between CD and
IBS (mean score: 79.42, P = 0.016). Mean perceived importance of
Table 1 . Baseline characteristics of celiac and non-celiac cohorts
CD HTN DM CHF ESRD GERD IBD IBS
Number of participants 341 75 69 41 18 61 67 37
Mean age, years (s.e.) 51.14 (0.81) 63.27 (1.37) 64.03 (1.31) 70.58 (2.07) 56.61 (4.27) 57.84 (1.62) 46.13 (1.90) 50.59 (0.81)
% Female, (c) 76.26 (0.02) 45.95 (0.06) 38.24 (0.06) 53.66 (0.08) 44.44 (0.12) 40.00 (0.06) 46.97 (0.06) 52.78 (0.36)
% Caucasian, (s.e.) 96.77 (0.01) 80.00 (0.05) 70.59 (0.06) 68.29 (0.07) 61.11 (0.12) 77.96 (0.05) 93.73 (0.03) 88.89 (0.05)
% Income
a < $ 75,000, (s.e.) 36.50 (0.03) 57.33 (0.06) 69.56 (0.06) 87.80 (0.05) 88.89 (0.08) 40.98 (0.06) 41.79 (0.06) 41.67 (0.08)
% No college education (s.e.) 26.71 (0.02) 48.00 (0.06) 56.52 (0.06) 56.10 (0.08) 66.67 (0.11) 42.62 (0.06) 28.36 (0.06) 30.56 (0.08)
Mean specialist visits per
year (s.e.)
0.75 (0.03) 0.48 (0.12) 1.65 (0.21) 2.34 (0.23) 4.39 (0.24) 1.26 (0.21) 2.51 (0.17) 1.82 (0.24)
Mean PCP visits per year (s.e.) 2.19 (0.08) 2.54 (0.16) 3.13 (0.18) 3.61 (0.24) 3.28 (0.36) 2.60 (0.16) 2.18 (0.18) 2.85 (0.28)
CD, celiac disease; CHF, congestive heart failure; DM, diabetes mellitus; ESRD, end-stage renal disease; GERD, gastroesophageal refl ux disease; HTN, hypertension; IBD,
infl ammatory bowel disease; IBS, irritable bowel syndrome.
a Income in US dollars.

THE RED SECTION
The American Journal of GASTROENTEROLOGY VOLUME 104 | XXX 2014 www.amjgastro.com
4
treatment was 90.14 for DM, 88.89 for HTN, 88.83 for CHF, 87.95
for IBD, and 86.49 for GERD.  e reported perceived importance
of treatment for CD of 93.80 was higher than all other groups in
aggregate (mean: 87.90, P < 0.001).
Overall and disease-speci c health . Participants with CD rated
disease-speci c health higher than any other group, with a mean
score of 81.61 (0: “ worst imaginable health state ” and 100: “ best
imaginable health state ” ).  is score was signi cantly higher
than that for ESRD (mean: 59.44, P = 0.003), IBS (mean: 65.20,
P = 0.003), CHF (mean score: 65.80, P = 0.002), and DM (mean:
71.22, P = 0.028). Disease-speci c health was higher than that
of the other chronic medical conditions as well, including HTN
(mean: 76.82), GERD (mean: 74.56), and IBD (mean: 74.34),
but did not reach statistical signi cance.  e reported disease-
speci c health for CD of 81.61 was higher than all other groups in
aggregate (mean: 71.72, P < 0.001).
Overall
CHF
DM
ESRD
HTN
IBS
IBD
GERD
CD
10 20 30 40 50 60 70 80
Visual analog scale
Figure 1 . Treatment burden of CD in comparison with non-celiac chronic
illnesses. * CD, celiac disease; CHF,congestive heart failure; DM, diabetes
mellitus; ESRD, end-stage renal disease; GERD,gastroesophageal
refl ux disease; HTN, hypertension; IBD,infl ammatory bowel disease;
IBS,irritable bowel syndrome. * * VAS (visual analog scale): score of
0 = very easy; score of 100 = very diffi cult. * * * Mean scores: CD 44.9
(s.d.: 30.9), GERD 21.3 (s.d.: 25.3), HTN 23.5 (s.d.: 25.7), IBD 31.9
(s.d.: 27.7), CHF 38.5 (s.d.: 31), IBS 40.4 (s.d.: 24.4), DM 41.7 (s.d.:
30.4), and ESRD 56.4 (s.d.: 31.9). * * * * On the basis of results of
post hoc multiple comparisons, treatment burden was statistically
signifi cant for the following groups: CD vs. GERD ( P value < 0.001), CD
vs. HTN ( P value < 0.001), GERD vs. ESRD ( P value 0.01), and GERD
vs. DM ( P value 0.04). Error bars represent 95 % confi dence intervals.
Overall
CHF
DM
ESRD
HTN
IBS
IBD
GERD
CD
60 8070 90 100
Visual analog scale
Figure 2 . Perceived importance of treatment of CD in comparison with
non-celiac chronic illnesses. * CD,celiac disease; CHF, congestive heart
failure; DM, diabetes mellitus; ESRD, end-stage renal disease; GERD,
gastroesophageal refl ux disease; HTN, hypertension; IBD, infl ammatory
bowel disease; IBS, irritable bowel syndrome. * * VAS (visual analog scale):
score of 0 = not important at all; score of 100 = very important. * * * Mean
scores: CD 93.8 (s.d.: 18.6), HTN 88.9 (s.d.: 20.4), GERD 86.5 (s.d.:
19.7), IBD 88.0 (s.d.: 19.9), CHF 88.8 (s.d.: 22.4), IBS 79.4 (s.d.: 25.5),
DM 90.1(s.d.: 20.4), and ESRD 94.7 (s.d.: 14.2). * * * * On the basis of
results of post hoc multiple comparisons, the importance of treatment
was statistically signifi cant for the following groups: CD vs. IBS ( P value
0.016). Error bars represent 95 % confi dence intervals.
Table 2 . Celiac patient factors associated with high GFD treatment burden
Univariate analysis Multivariate analysis
β coeffi cient 95 % CI β coeffi cient 95 % CI
Time limitations 14.21 6.09, 22.33 10.41 2.51, 18.31
Eating outside the home 12.64 3.82, 21.46 10.13 1.44, 18.82
Concern regarding food cost 11.41 4.63, 18.18 8.91 1.94, 15.88
No college education 10.78 3.48, 18.09 9.73 2.33, 17.14
Hospitalization in the past year 9.01 2.62, 15.39 NS NS
Poor GFD adherence 2.42 1.54, 3.31 1.77 0.82, 2.72
Increased celiac symptoms 0.61 0.27, 0.96 NS NS
Income
a > $ 200,000 NS NS 11.51 2.08, 20.94
CI, confi dence intervals; NS, not signifi cant.
a Income in US dollars.

THE RED SECTION
© 2014 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
5
compared with 31.71 ( P = 0.89) in those not on biologic therapy.
Disease-speci c health and overall health were 76.40 and 79.80,
respectively, compared with 73.96 ( P = 0.778) and 76.11 ( P = 0.576)
for those not on biologic therapy. When compared with IBD
participants not on biologics, celiac participants reported greater
disease-speci c health (73.96 vs. 81.61, P = 0.036), treatment burden
(mean di erence: 31.71 vs. 44.90, P = 0.002), and perceived impor-
tance of treatment (mean di erence: 87.26 vs. 93.80, P = 0.027).
No di erence was observed between IBD participants on biologic
therapy and celiac participants for disease-speci c health (76.40
vs. 81.61, P = 0.479), treatment burden (33.00 vs. 44.90, P = 0.243),
and importance of treatment (92.11 vs. 93.80, P = 0.808).
Similarly, participants with CD rated overall health higher than
any other group, with a mean score of 77.97. Overall health was
signi cantly greater for CD compared with ESRD (mean: 55.44,
P = 0.001) and CHF (mean: 57.80, P < 0.001), but it did not reach
signi cance for HTN (mean: 72.44), GERD (mean: 71.62), IBD
(mean: 76.66), or IBS (mean: 70.54).  e reported overall health
for CD of 77.97 was higher than all other groups in aggregate
(mean: 69.93, P < 0.001).
In subgroup analysis, we further characterized IBD patients on
the basis of the use of biologic therapy and DM patients on the basis
of the use of insulin. Fi een percent of IBD participants reported
taking biologics, and in this group treatment burden was 33.00
Overall
CHF
DM
ESRD
HTN
IBS
IBD
GERD
CD
50 60 8070 90
Visual analog scale
Figure 3 . Perceived disease-specifi c health of CD in comparison with
non-celiac chronic illnesses. * CD, celiac disease; CHF, congestive heart
failure; DM, diabetes mellitus; ESRD, end-stage renal disease; GERD,
gastroesophageal refl ux disease; HTN, hypertension; IBD, infl ammatory
bowel disease; IBS, irritable bowel syndrome. * * VAS (visual analog scale):
score of 0 = best imaginable health, score of 100 = worst imaginable health.
* * * Mean scores: CD 81.6 (s.d.: 18), GERD 74.6 (s.d.: 18.4), HTN 76.8
(s.d.: 18.4), IBD 74.3 (s.d.: 24.9), CHF 65.8 (s.d.: 19.4), IBS 65.2 (s.d.:
20.7), DM 71.2 (s.d.: 22.7), and ESRD 59.4 (s.d.: 22.2). * * * * On the basis
of results of post hoc multiple comparisons, disease-specifi c health was
statistically signifi cant for the following groups: CD vs. IBS ( P value 0.003),
CD vs. ESRD ( P value 0.003), CD vs. DM ( P value 0.028), and CD vs. CHF
( P value 0.002). Error bars represent 95 % confi dence intervals.
Overall
CHF
DM
ESRD
HTN
IBS
IBD
GERD
CD
50 60 8070
Visual analog scale
Figure 4 . Perceived overall health state of CD in comparison with non-celiac
chronic illnesses. * CD, celiac disease; CHF, congestive heart failure; DM,
diabetes mellitus; ESRD, end-stage renal disease; GERD, gastroesophageal
refl ux disease; HTN, hypertension; IBD, infl ammatory bowel disease; IBS,
irritable bowel syndrome. * * VAS (visual analog scale): score of 0 = best
imaginable health, score of 100 = worst imaginable health. * * * Mean scores:
CD 78 (s.d.: 16.7), GERD 71.6 (s.d.: 21.6), HTN 72.4 (s.d.: 20.9), IBD 76.7
(s.d.: 19.1), CHF 57.8 (s.d.: 21.2), IBS 70.5 (s.d.: 18.1), DM 69.9 (s.d.: 23),
and ESRD 55.4 (s.d.: 17.8). * * * * On the basis of results of post hoc multiple
comparisons, overall health was statistically signifi cant for the following
groups: CD vs. ESRD ( P value 0.001), CD vs. CHF ( P value < 0.001), IBD
vs. ESRD ( P value 0.014), IBD vs. CHF ( P value 0.001), and HTN vs. CHF
( P value 0.029). Error bars represent 95 % confi dence intervals.
Table 3 . Celiac patient factors associated with poor adherence to GFD
Univariate analysis Multivariate analysis
β coeffi cient 95 % CI β coeffi cient 95 % CI
Income
a < $ 200,000 1.90 0.04, 3.76 0.92 (0.02, 1.81)
Unemployed 1.43 (0.002, 2.87) NS NS
Worse symptoms 0.17 (0.13, 0.20) 0.19 (0.15, 0.22)
Lower perceived treatment importance 0.07 (0.05, 0.09) 0.09 (0.07, 0.11)
Greater treatment burden 0.03 (0.02, 0.04) NS NS
CI, confi dence intervals; GFD, gluten-free diet; NS, not signifi cant.
a Income in US dollars.

THE RED SECTION
The American Journal of GASTROENTEROLOGY VOLUME 104 | XXX 2014 www.amjgastro.com
6
 irty-two percent of diabetic participants reported taking
insulin. Treatment burden (45.00 vs. 40.43, P = 0.583), disease-
speci c health (70.82 vs. 71.41, P = 0.920), and overall health
(67.50 vs. 71.06, P = 0.554) did not di er for diabetics on insu-
lin compared with those not on insulin. When compared with
diabetics not on insulin, celiac participants reported greater
disease-speci c health (71.41 vs. 81.61, P = 0.004). Treatment
burden (40.43 vs. 44.90, P = 0.346) and perceived importance
of treatment (87.72 vs. 93.80, P = 0.101) did not di er between
diabetics not on insulin and celiac participants. Disease-speci c
health (70.82 vs. 81.61, P = 0.060), treatment burden (45.00 vs.
44.90, P = 0.989), and treatment importance (95.32 vs. 93.80,
P = 0.374) did not di er between diabetics on insulin and celiac
participants.
DISCUSSION
Adherence to a GFD is the only treatment for CD, and failure to
adequately treat is associated with morbidity. However, GFD adher-
ence remains poor (40,48), and studies examining factors that
a ect treatment burden and GFD adherence are limited ( 36,38,49 ).
We sought to evaluate predictors of adherence and treatment bur-
den, and to compare treatment burden in CD with other chronic
diseases.
Our results demonstrate that celiac participants report a
remarkably high treatment burden. In our study, celiac partici-
pants reported greater treatment burden than those with HTN
and GERD and comparable to participants with CHF and ESRD.
 is underscores the di culty of following the GFD and puts the
high treatment burden of CD into context for practitioners who
o en have more experience with other diseases. Greater di culty
with treatment implies a need for nondietary interventions, as
well as the need for patients to regularly follow-up with clini-
cians, dieticians, and other allied health professionals, whereas
our data and others ’ ( 50 ) suggest that patients with CD actually
see physicians less o en than those with other chronic medical
conditions.
Our results also show that treatment burden is a predictor of
poor adherence to a GFD.  ose who report higher burden of
following a GFD were more likely to have poor adherence to a
GFD. An exception to this was in participants with high reported
household income.  ese individuals were likely to report high
treatment burden but were able to overcome this and demon-
strate greater adherence to the GFD. Although our study was not
designed to evaluate this relationship in detail, we hypothesize that
low education increases burden owing to di culty in managing
the complexity of the GFD, whereas high income increases bur-
den owing to di culty in following the diet during frequent travel
and social outings. Similar to prior studies ( 41 ), celiac participants
with poor GFD adherence had greater severity of symptoms likely
because of gluten exposure. Although the majority of participants
reported overall high perceived importance of the GFD, the  nd-
ing that participants who felt the GFD was not very important to
their health had poor adherence is logical and re ects the internal
validity of the survey measures.
It is encouraging to note that despite having reported quite a
high treatment burden in CD when compared with other chronic
medical conditions, participants with CD also reported a high
disease-speci c and overall health. It is notable that CD patients
reported greatest disease-speci c health, yet reported higher treat-
ment burden compared with all other diseases except ESRD.  e
high burden of treatment may also be addressed by focusing on
other alternative treatments such as novel medical therapy or even
complementary alternative medicine ( 51 ).
Despite the use of validated measures and the relatively large
size, we recognize a number of limitations.  e data were collected
from a dedicated CD center at a major teaching hospital, which
potentially limits the generalizability of the  ndings. In addition,
there was a substantial nonresponse rate that may bias results if
respondents di er signi cantly from nonrespondents; however, in
the CD cohort, respondent demographics were not signi cantly
di erent from nonrespondents. ( Supplementary Table S1 online)
In addition, the response rate of 45 % for the CD patients compared
with 29 % for non-CD patients raises the possibility of di erential
selection between these groups, which could bias results. Finally,
although the VAS has been widely used and validated in other dis-
ease states ( 42 – 46 ), it is possible that the VAS may not measure the
full spectrum of treatment burden, importance of treatment, and
health state. In addition, participants were asked to rate treatment
burden and health state without knowledge of the other disease
states with which they were compared. Although there is value in
asking patients to rate their health compared with another distinct
health state, it is di cult for individuals to gauge the impact of
conditions they do not have. For this reason, we chose to use the
Table 4 . Celiac patient factors associated with higher perceived importance of following GFD
Univariate analysis Multivariate analysis
β coeffi cient 95 % CI β coeffi cient 95 % CI
Eating out 6.06 0.74, 11.38 NS NS
Female gender 4.60 0.01, 9.19 NS NS
Greater adherence 1.92 1.40, 2.43 2.71 2.19, 3.23
Worse symptoms 0.24 0.03, 0.45 0.70 0.51, 0.89
CI, confi dence intervals; GFD, gluten-free diet; NS, not signifi cant.

THE RED SECTION
© 2014 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
7
1 3 . N a c h m a n F , M a u r i n o E , V a z q u e z H et al. Quality of life in celiac disease
patients: prospective analysis on the importance of clinical severity at
diagnosis and the impact of treatment . Dig Liver Dis 2009 ; 41 : 15 – 25 .
1 4 . N a c h m a n F , d e l C a m p o M P , G o n z a l e z A et al. Long-term deterioration
of quality of life in adult patients with celiac disease is associated with
treatment noncompliance . Dig Liver Dis 2010 ; 42 : 685 – 91 .
15 . Mustalahti K , Lohinierni S , Laippala P et al. Improvement of the quality
of life of silent celiac disease patients during the gluten-free diet warrents
screening . Gastroenterology 2000 ; 118 : A369 .
1 6 . V a s q u e z H , M a z u r e R , G o n z a l e z D et al. Risk of fractures in celiac disease
patients: a cross-sectional, case-control study . Am J Gastroenterol 2000 ;
95 : 183 – 9 .
17 . Valdimarsson T , Lofman O , Toss G et al. Reversal of osteopenia with diet
in adult coeliac disease . Gut 1996 ; 38 : 322 – 7 .
1 8 . S a t e g n a - G u i d e t t i C , G r o s s o S B , G r o s s o S et al.  e e ects of 1-year gluten
withdrawal on bone mass, bone metabolism and nutritional status in
newly-diagnosed adult coeliac disease patients . Aliment Pharmacol  er
2000 ; 14 : 35 – 43 .
1 9 . M u s t a l a h t i K , C o l l i n P , S i e v a n e n H et al. Osteopenia in patients with
clinically silent coeliac disease warrants screening . Lancet 1999 ; 354 : 744 – 5 .
2 0 . M e y e r D , S t a v r o p o l o u s S , D i a m o n d B et al. Osteoporosis in a north american
adult population with celiac disease . Am J Gastroenterol 2001 ; 96 : 112 – 9 .
21 . Ciacci C , Maurelli L , Klain M et al. E  ects of dietary treatment on bone
mineral density in adults with celiac disease: factors predicting response .
Am J Gastroenterol 1997 ; 92 : 992 – 6 .
2 2 . S i n i s c a l c h i M , I o v i n o P , T o r t o r a R et al. Fatigue in adult coeliac disease .
Aliment Pharmacol  er 2005 ; 22 : 489 – 94 .
23 . Sher KS , Mayberry JF . Female fertility, obstetric and gynaecological history
in coeliac disease: a case control study . Acta Paediatr 1996 ; 412 : 76 – 7 .
24 . Rujner J . Age at menarche in girls with celiac disease . Ginekol Polska
1999 ; 70 : 359 – 62 .
2 5 . F e r g u s o n R , H o l m e s G K , C o o k e W T . C o e l i a c d i s e a s e , f e r t i l i t y , a n d
pregnancy . Scand J Gastroenterol 1982 ; 17 : 65 – 8 .
2 6 . B o n a G , M a r i n e l l o D , O d e r d a G . M e c h a n i s m s o f a b n o r m a l p u b e r t y i n
coeliac disease . Horm Res 2002 ; 57 (Suppl 2) : 63 – 5 .
2 7 . N o r g a r d B , F o n a g e r K , S o r e n s e n H T et al. Birth outcomes of women with
celiac disease: a nationwide historical cohort study . Am J Gastroenterol
1999 ; 94 : 2435 – 40 .
2 8 . L u d v i g s s o n J F , M o n t g o m e r y S M , E k b o m A . C e l i a c d i s e a s e a n d r i s k o f
adverse fetal outcome: a population-based cohort study . Gastroenterology
2005 ; 129 : 454 – 63 .
29 . Ciacci C , Cirillo M , Auriemma G et al. Celiac disease and pregnancy
outcome . Am J Gastroenterol 1996 ; 91 : 718 – 22 .
3 0 . L e b w o h l B , G r a n a t h F , E k b o m A et al. Mucosal healing and risk for lym-
phoproliferative malignancy in celiac disease: a population-based cohort
study . Ann Intern Med 2013 ; 159 : 169 – 75 .
3 1 . L u d v i g s s o n J F . M o rtality and malignancy in celiac disease . Gastrointest
Endosc Clin N Am 2012 ; 22 : 705 – 22 .
32 . Stevens L , Rashid M . Gluten-free and regular foods: a cost comparison .
Can J Diet Pract Res 2008 ; 69 : 147 – 50 .
33 . Singh J , Whelan K . Limited availability and higher cost of gluten-free foods .
J Hum Nutr Diet 2011 ; 24 : 479 – 86 .
3 4 . L e e A R , N g D L , Z i v i n J et al. Economic burden of a gluten-free diet . J Hum
Nutr Diet 2007 ; 20 : 423 – 30 .
3 5 . W h i t a k e r J K , W e s t J , H o l m e s G K et al. Patient perceptions of the burden
of coeliac disease and its treatment in the UK . Aliment Pharmacol  er
2009 ; 29 : 1131 – 6 .
3 6 . E d w a r d s G e o r g e J B , L e  e r D A , D e n n i s M D et al. Psychological correlates
of gluten-free diet adherence in adults with celiac disease . J Clin Gastro-
enterol 2009 ; 43 : 301 – 6 .
37 . DiMatteo MR . Variations in patients ’ adherence to medical recommenda-
tions: a quantitative review of 50 years of research . Med Care 2004 ; 42 : 200 – 9 .
3 8 . L e  e r D A , E d w a r d s - G e o r g e J , D e n n i s M et al. Factors that in uence
adherence to a gluten-free diet in adults with celiac disease . Dig Dis Sci
2008 ; 53 : 1573 – 81 .
3 9 . L e  er D . Celiac disease diagnosis and management: a 46-year-old woman
with anemia . JAMA 2011 ; 306 : 1582 – 92 .
4 0 . H a l l N J , R u b i n G , C h a r n o c k A . S y s t e m a t i c r e v i e w : a d h e r e n c e t o a g l u t e n -
free diet in adult patients with coeliac disease . Aliment Pharmacol  er
2009 ; 30 : 315 – 30 .
4 1 . L e  e r D A , D e n n i s M , E d w a r d s G e o r g e J B et al. A simple validated gluten-
free diet adherence survey for adults with celiac disease . Clin Gastroenterol
Hepatol 2009 ; 7 : 530 – 6 , 6.e1 – 2 .
common and well-validated anchors such as of “ best imaginable
health ” and “ worst imaginable health ” .
In conclusion, participants with CD report a remarkably high
treatment burden similar to participants with ESRD and higher
than many other chronic medical conditions. Conversely, CD
participants in general reported high disease-speci c health state,
which suggests that despite good long-term health outcomes CD
patients struggle with treatment.  is study underscores both the
limitations of the GFD as lone treatment and the need for attention
to patient perceptions of recommended therapies ( 51 ). As the CD
population expands and new therapies are proposed, attention to
burden of treatment and disease will be vital for providing optimal
care for this population.
CONFLICT OF INTEREST
Guarantor of the article : Daniel A. Le er, MD, MS.
Speci c author contributions: Study design and execution, analysis
and interpretation of data, dra ing of the manuscript, and critical
revision of the manuscript for important intellectual content of the
manuscript: Sveta Shah and Mona Akbari; study execution: Rohini
Vanga, Joshua Hansen,  immaiah  eethira, and Sohaib Tariq;
study design and execution, interpretation of data, and critical
revision of the manuscript for important intellectual content of
the manuscript: Ciar á n P. Kelly; study design, execution, and
enrollment: Melinda Dennis; study concept, design and execution,
interpretation of data, and critical revision of the manuscript for
important intellectual content of the manuscript: Daniel A. Le er.
Financial support:  is work was supported by Shire  erapeutics,
Prometheus Laboratories, Alba Pharmaceuticals, and Alvine
 erapeutics (D.A.L.).
Potential competing interests: N o n e .
REFERENCES
1 . L o h i S , M u s t a l a h t i K , K a u k i n e n K et al. Increasing prevalence of coeliac
disease over time . Aliment Pharmacol  er 2007 ; 26 : 1217 – 25 .
2 . F a s a n o A , B e r t i I , G e r a r d u z z i T et al. Prevalence of celiac disease in at-risk
and not-at-risk groups in the United States: a large multicenter study .
Arch Int Med 2003 ; 163 : 286 – 92 .
3 . Farrell RJ , Kelly CP . Celiac sprue . N Engl J Med 2002 ; 346 : 180 – 8 .
4 . Rubio-Tapia A , Ludvigsson JF , Brantner TL et al.  e prevalence of celiac
disease in the United States . Am J Gastroenterol 2012 ; 107 : 1538 – 44 ; quiz 7, 45 .
5 . G u j r a l N , F r e e m a n H J ,  omson AB . Celiac disease: prevalence, diagnosis,
pathogenesis and treatment . World J Gastroenterol 2012 ; 18 : 6036 – 59 .
6 . V a h e d i K , M a s c a r t F , M a r y J Y et al. Reliability of antitransglutaminase
antibodies as predictors of gluten-free diet compliance in adult celiac
disease . Am J Gastroenterol 2003 ; 98 : 1079 – 87 .
7 . Hogberg L , Grodzinsky E , Stenhammar L . Better dietary compliance in
patients with coeliac disease diagnosed in early childhood . Scand J
Gastroenterol 2003 ; 38 : 751 – 4 .
8 . B a r d e l l a M T , M o l t e n i N , P r a m p o l i n i L et al. Need for follow up in coeliac
disease . Arch Dis Child 1994 ; 70 : 211 – 3 .
9 . L e  e r D A , D e n n i s M , H y e t t B et al. Etiologies and predictors of diagnosis
in nonresponsive celiac disease . Clin Gastroenterol Hepatol 2007 ; 5 :
4 4 5 – 5 0 .
1 0 . L u d v i g s s o n J F , M o n t g o m e r y S M , E k b o m A et al. Small-intestinal histo-
pathology and mortality risk in celiac disease . JAMA 2009 ; 302 : 1171 – 8 .
1 1 . L e  er D , Kelly CP . Celiac disease: what the last few years have taught us .
In: Howden CW (ed). Advances in Digestive Diseases . AGA Institute Press:
Bethesda, MD, USA , 2007 , pp. 49 – 56 .
1 2 . T o n t i n i G E , R o n d o n o t t i E , S a l a d i n o V et al. Impact of gluten withdrawal on
health-related quality of life in celiac subjects: an observational case-control
study . Digestion 2010 ; 82 : 221 – 8 .

THE RED SECTION
The American Journal of GASTROENTEROLOGY VOLUME 104 | XXX 2014 www.amjgastro.com
8
4 2 . W i l c o x A R , D r a g n e v M C , D a r c e y C J et al. A new tool to measure the
burden of Crohn’s disease and its treatment: do patient and physician
perceptions match? In amm Bowel Dis 2010 ; 16 : 645 – 50 .
43 . Liu JY , Woloshin S , Laycock WS et al. Symptoms and treatment burden of
gastroesophageal re ux disease: validating the GERD assessment scales .
Arch Int Med 2004 ; 164 : 2058 – 64 .
4 4 . J a n s s e n M F , L u b e t k i n E I , S e k h o b o J P et al.  e use of the EQ-5D preference-
based health status measure in adults with Type 2 diabetes mellitus .
Diab Med 2011 ; 28 : 395 – 413 .
4 5 . E k m a n I , G r a n g e r B , S w e d b e r g K et al. Measuring shortness of breath in
heart failure (SOB-HF): development and validation of a new dyspnoea
assessment tool . Eur J Heart Fail 2011 ; 13 : 838 – 45 .
46 . Bushnell DM , Martin ML , Ricci JF et al. Performance of the EQ-5D in
patients with irritable bowel syndrome . Value Health 2006 ; 9 : 90 – 7 .
4 7 . L e  e r D A , D e n n i s M , E d w a r d s G e o r g e J et al. A validated disease-speci c
symptom index for adults with celiac disease . Clin Gastroenterol Hepatol
2009 ; 7 : 1328 – 34 , 34.e1 – 3 .
48 . Pietzak MM . Follow-up of patients with celiac disease: achieving compli-
ance with treatment . Gastroenterology 2005 ; 128 (4 Suppl 1) : S135 – 41 .
49 . Ciacci C , Cirillo M , Cavallaro R et al. Long-term follow-up of celiac
adults on gluten-free diet: prevalence and correlates of intestinal damage .
Digestion 2002 ; 66 : 178 – 85 .
5 0 . H e r m a n M L , R u b i o - Ta p i a A , L a h r B D et al. Patients with celiac disease
are not followed up adequately . Clin Gastroenterol Hepatol 2012 ; 10 :
893 – 899. e1 .
5 1 . A z i z I , E v a n s K E , P a p a g e o r g i o u V et al. Are patients with coeliac disease
seeking alternative therapies to a gluten-free diet? J Gastrointestin Liver Dis
2011 ; 20 : 27 – 31 .