Beruflich Dokumente
Kultur Dokumente
Pertti Aro, Nicholas J. Talley, Sven-Erik Johansson, Lars Agrus, Jukka Ronkainen
PII: S0016-5085(15)00156-0
DOI: 10.1053/j.gastro.2015.01.039
Reference: YGAST 59591
Please cite this article as: Aro P, Talley NJ, Johansson S-E, Agrus L, Ronkainen J, Anxiety is Linked to
New-onset Dyspepsia in the Swedish Population A 10 year follow-up Study, Gastroenterology (2015),
doi: 10.1053/j.gastro.2015.01.039.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
All studies published in Gastroenterology are embargoed until 3PM ET of the day they are published as
corrected proofs on-line. Studies cannot be publicized as accepted manuscripts or uncorrected proofs.
1
ACCEPTED MANUSCRIPT
PT
Pertti Aro1, Nicholas J Talley2, Sven-Erik Johansson1, Lars Agrus1, Jukka Ronkainen1,3,4
1
RI
Centre for Family Medicine, Karolinska Institutet, Stockholm, Sweden,
2
Faculty of Health and Medicine, University of Newcastle, Callaghan, NSW, Australia,
SC
3
Institute of Health Sciences, Medical Faculty, University of Oulu, Finland,
4
Primary Health Care Centre, Tornio, Finland,
U
Abbreviations: 99% CI 99% Confidence Interval, EGD - esophagogastroduodenoscopy,
AN
FD - Functional dyspepsia, H. pylori - Helicobacter pylori, IBS irritable bowel syndrome,
NSAID - Non Steroidal Anti Inflammatory Drug, OR odds ratio, PPI - Proton Pump
D
Inhibitor
TE
Address for correspondence: Dr. Pertti Aro, Taimenkuja 4, SF-95410, Tornio Finland. Tel:
Grant support
This study was supported in part by the Swedish Research Council, the Swedish Society of
C
Medicine, Norrbotten County Council (Sweden), Astra Zeneca R&D (Sweden), Orion
AC
Research Foundation (Finland), the Finnish Medical Foundation (Finland) and Vappu and
The study sponsors did not have role in the study design in the collection, analysis, and
interpretation of data.
Disclosures
2
ACCEPTED MANUSCRIPT
The authors have nothing to disclose
Author Contributions
Pertti Aro: Constructing the study, performing the study and collecting the data, analysing the
Nicholas Talley: Constructing the study, supervising the study, writing and revising the article
PT
Sven-Erik Johansson: Analysing the data, revising the article
RI
Lars Agrus: Constructing the study, supervising the study, writing and revising the article
Jukka Ronkainen: Constructing the study, performing the study and collecting the data,
SC
writing and revising the article.
U
AN
M
D
TE
C EP
AC
3
ACCEPTED MANUSCRIPT
ABSTRACT
Backgroud & Aims: Functional dyspepsia (FD) is associated with anxiety but it is not clear
if one causes the other. We investigated whether anxiety and depression precede the onset of
FD (based on the modified Rome III criteria) and gastroesophageal reflux symptoms (GERS)
PT
Methods: Participants from the Kalixanda study (n=3000), randomly selected from the
RI
national population register of Sweden, were given the validated abdominal symptom
questionnaire (ASQ) 19982001; 1000 of these participants were then selected randomly to
SC
undergo esophagogastroduodenoscopy and were given the ASQ along with hospital anxiety
and depression scale (HADS) questionnaire. All eligible subjects who underwent endoscopy
U
(n=887) were invited to a follow-up study in JuneAugust 2010 and were given the same
AN
questionnaires. Data were analyzed by logistic regression.
M
Results: Of the 703 subjects who completed the follow-up questionnaires (79.3%); 110 were
these were new cases of FD. GERS without organic disease was reported by 273 individuals
TE
(38.8%) at baseline and by 280 at follow-up examination (39.8%); 93 cases were new. Major
anxiety was associated with FD at follow-up (odds ratio [OD]=6.30; 99% confidence interval
EP
[CI], 1.6424.16). Anxiety was associated with post-prandial distress syndrome at baseline
C
(OR=4.83; 99% CI, 1.2418.76) and at the follow-up examination (OR=8.12; 99% CI, 2.13
AC
30.85) but not with epigastric pain syndrome. Anxiety at baseline was associated with new-
onset FD at the follow-up examination (OR=7.61; 99% CI, 1.2147.73), but not with GERS.
Conclusions: In a study of the Swedish population, anxiety at baseline, but not depression,
increased risk for development of FD 7.6-fold in the next 10 years. Anxiety did not affect risk
for GERS.
Functional dyspepsia (FD) and gastroesophageal reflux symptoms (GERS) are common and
costly disorders in the adult population.(1-4) The prevalence of FD in different studies depends
on the definition used(5-7) but the prevalence of FD in the western world defined according to
the Rome III definition is between 11% and 20%(4, 8) whereas the prevalence of
PT
gastroesophageal reflux symptoms (GERS) once a week or more often is approximately
RI
20%.(9, 10)
SC
Psychological distress has been reported to be linked with dyspepsia(11-14) as has
U
reported to cause emotional distress.(16) We have earlier described the association of FD with
AN
anxiety(4) but longitudinal follow-up studies in the general population applying the Rome III
M
definition are largely unavailable.(17) There is one 5-year follow-up study on FD patients
applying the Rome II definition showing an association of anxiety with FD(18) and another
D
of anxiety and depression with GERS(20) whereas Eslick and others in Australia did not find
AC
determine if there is an association of psychological factors with GERS are not available.
Any link between GI symptoms and mental distress could be confounded by other
environmental risk factors such as smoking or alcohol use.(22-25) Smoking is linked to post
infectious FD(26) and to GERS(21) but whether this environmental factor drives new onset of
5
ACCEPTED MANUSCRIPT
symptoms is unknown. An association of non-steroidal anti-inflamatory drugs (NSAIDs) with
some NSAIDs and for high dose of any NSAID(27) whereas an Italian population-based
endoscopic study failed to find any association of NSAIDs with FD.(8) There are also
inconsistent reports whether obesity and GI symptoms are linked.(28-30) There was no relation
PT
between body mass index and GERS in a Swedish population-based study(28) whereas others
have found that being overweight and obese are independent risk factors of GERS.(29)
RI
SC
The aim of this study was to define whether there may be a causal relation between anxiety
and depression with functional dyspepsia and/or GERS. We also aimed to study the overlap
U
of dyspepsia and GERS, and predictors for new onset of symptoms. We hypothesized that
AN
anxiety and/or depression drive FD and/or GERS and predict the new onset of dyspepsia.
M
D
TE
C EP
AC
6
ACCEPTED MANUSCRIPT
METHODS
The baseline data for the ten year follow-up study were gathered from a representative
random sample (n=3,000) of the general population in two Swedish communities between
1998 and 2001. At the baseline the participants were randomly selected from the national
population register which covers all inhabitants in Sweden and surveyed by a validated and
PT
widely used abdominal symptom questionnaire (ASQ).(31, 32) One thousand of them were
RI
randomly selected for esophagogastroduodenoscopy and for survey by a more comprehensive
ASQ assessing troublesome gastrointestinal symptoms and by the hospital anxiety and
SC
depression scale (HADS)(33) to define depression and anxiety. Organic reasons for dyspepsia
U
Kalixanda study.(34) All eligible from this endoscoped cohort (n=887) were invited to a
AN
follow-up study in June-August 2010 with the same postal questionnaires ten years after the
M
primary investigation (Figure 1). During these years the subjects could consult and be treated
Aside from the individual symptoms from the questionnaires, the participants were asked
about their level of education, gastrointestinal (GI) medication and weight as well as their use
EP
The baseline study was approved by the Ume University ethics committee and the follow-up
study by the ethical approval committee of the Karolinska Institutet and conducted in
Endoscopy at baseline
7
ACCEPTED MANUSCRIPT
The upper esophagogastroduodenoscopies (EGDs) were undertaken by three experienced
endoscopists, in the two clinics (Kalix and Haparanda), which gave sole medical cover to the
area. Internal validity was assessed by means of consensus sessions before the initiation of the
basic study. The endoscopists had been participating in regular quality assessment programs
over several years. The endoscopists were unaware of the symptoms of the subjects before
PT
endoscopy. Only topical anaesthesia was used.(34)
RI
Definitions of dyspepsia, gastroesophageal reflux symptoms and irritable bowel syndrome
SC
Functional dyspepsia (FD) was defined based on the Rome III definition: weekly bothersome
postprandial fullness or early satiation, or epigastric pain and/or epigastric burning without
U
findings of esophagitis, peptic ulcer, celiac disease or cancer and no evidence of other
AN
structural disease at endoscopy that was likely to explain the symptoms. FD, according to the
M
2. Epigastric Pain Syndrome (EPS) consisting of pain or burning localized to the epigastric
area and not generalized or localized to other abdominal or chest regions, and not
EP
relieved by defecation. Overlap between PDS and EPS was allowed in line with the
(GERS) or irritable bowel syndrome (IBS) did not exclude the diagnosis of FD.(7)
AC
The ASQ was designed before the Rome era but the questionnaire has been updated to meet
both the Rome II and III criteria and it measures all of the Rome III criteria aside from the
acid regurgitation over the past three months. The presence of FD or IBS did not exclude the
Irritable bowel syndrome (IBS) was defined as troublesome abdominal pain or discomfort
PT
located at any site plus concomitant bowel habit disturbances (constipation, diarrhea, or
RI
alternating constipation and diarrhea). This simple definition has been used previously and
shown to produce results reasonably concordant with the Rome I criteria for IBS.(38-40)
SC
Definition of anxiety and depression
U
A HADS score from 8 to less than 11 was used to define suspected anxiety and depression
AN
and 11 or more was used as cut off level for both clinically relevant (major) anxiety and
depression.(33)
M
D
Definiton of obesity
TE
Body mass index (BMI) was calculated according to guidelines from WHO and it was
categorized as normal with BMI <25, overweight with a BMI between 25 and <30 and obesity
EP
Covariates
All individuals were thoroughly interviewed face to face at the baseline study about their
smoking habits, use of alcohol and use of medication including use of non-steroidal anti-
10-year follow-up study. The question about the dose of alcohol/week was also included. The
cut-off level for high consumption of alcohol was set at 100 grams absolute alcohol or
more/week.
PT
Statistics
RI
Logistic regression adjusting for age and gender was used in cross-sectional analyses both at
SC
overweight, obesity), categorized anxiety (no anxiety, suspected anxiety and major anxiety),
categorized depression (no depression, suspected depression and major depression), use of
U
proton pump inhibitors (PPIs), high consumption of alcohol, smoking, Helicobacter pylori (H.
AN
pylori) infection and use of NSAIDs and/or aspirin with FD, FDs subgroups (PDS, EPS and
M
epigastric burning), GERS and IBS. Analyses were performed separately for different
gastrointestinal symptoms.
D
Predictors for change in presentation of FD, PDS and EPS, GERS and IBS were analyzed in
TE
separate analyses by logistic regression adjusting for age and gender and including categorized
BMI, categorized anxiety, categorized depression, PPIs, smoking, H. pylori infection and use
EP
of NSAIDs and/or aspirin in the model. Unadjusted odds ratios (crude ORs) for association of
anxiety and depression with FD and GERS are also presented. Due to non-linearity, age was
C
AC
dichotomized with cut off at 60 years of age. Variables were chosen according to our
hypothesis and due to their clinical significance. Only statistically significant variables in
univariate anlysis were introduced into the basic model adjusted for age and gender. Model
selection was performed by dropping out statistically non-significant independent variables one
by one and the goodness of-fit of the new model was judged after each variable was removed
using the likelihood-ratio test. The reduced model was judged to be better if the p value of the
10
ACCEPTED MANUSCRIPT
likelihood-ratio test was >0.05. The goodness-of fit test was then performed after the chosen
model (Pearssons 2 test) and if the p-value was >0.05 the model was judged to fit the data.(42)
Fischers exact test was applied when number of observations was less than 10. The T-test was
used to analyze significances of differences in mean anxiety scores at baseline and follow-up in
PT
P-values were all two tailed and the alpha level of significance was set at 0.01. The results are
RI
shown as odds ratios (OR) with a 99% CI. The Intercooled STATA 9 program was used for
analyses.(43)
U SC
AN
M
D
TE
C EP
AC
11
ACCEPTED MANUSCRIPT
RESULTS
Of the original study population (n=3,000), 2,860 were eligible for inclusion. The overall
response rate was 74.2 % (n=2,122) at baseline and 1,364 responders to the ASQ eligible for
upper endoscopy were invited in random order to EGD with a response rate of 73.3%
(n=1,000). Ten years after the baseline study, 887 subjects were eligible for the follow-up
PT
study: 113 individuals had deceased, moved away from Sweden, could not be reached for other
RI
reasons or had developed a serious mental or physical illness during the 10 year period so that
they could not answer the questionnaires. Overall 703 out of 887 subjects answered the postal
SC
questionnaires, a response rate of 79.3%. BMI could be defined in 575 individuals (81.8% of
the responders).
U
AN
Mean age at the baseline was 53.0 years and at the follow-up 63.2 years, and the proportion of
M
women was 51.9 % in this follow-up cohort. Mean BMI was 26.3 both at the baseline and at
the follow up. Of those who were obese at the baseline (n=90), 74.4% were still obese (n=67).
D
At baseline 14.4% were smokers and the respective figure at follow-up was 11.6%.
TE
At baseline 12.2% used 100 g/week or more alcohol compared to 7.2% at follow-up.
Treatment with PPIs over the 3 months prior to the baseline study was reported by 5.1%
EP
(n=36) and at the 10 year follow-up period such therapy was reported by 21.6% (n=145).
C
AC
FD was identified in 110 of these 703 individuals (15.6%) at the baseline study and by 93
individuals at follow-up (13.3%); these groups overlapped in 45 (6.4%) individuals. EPS was
reported by 35 (5.0%) and PDS by 87 (12.4%) individuals at baseline and there was overlap in
12 individuals. There were 21 EPS (3.0%) and 79 PDS cases (11.2%) at follow-up overlapping
In cross-sectional analysis, major anxiety, but not depression was associated with FD only at
follow-up. (Table 1) Smoking, being overweight or obese, H. pylori infection, use of alcohol
and use of NSAIDs or aspirin were not associated with FD. However, PDS was associated with
anxiety both at baseline (crude OR= 4.83; 99% CI 1.26-18.52, adjusted OR=4.83; 99% CI
PT
1.24-18.76) and at follow-up (crude OR= 9.65; 99% CI 3.87-24.06, adjusted OR=8.12; 99% CI
RI
2.13-30.85) but EPS and separately epigastric burning were not.
SC
GERS were reported by 273 individuals (38.8%) at baseline and by 280 individuals (39.8%) at
follow-up and the groups overlapped in 187 individuals. Suspected and major anxiety were not
U
associated with GERS at baseline. Major anxiety was associated with GERS at follow-up both
AN
in unadjusted and adjusted analyses (Table 1). Being overweight was not associated with
M
GERS at baseline (crude OR=1.19; 99% CI 0.71-1.99, adjusted OR= 1.35; 99% CI 0.84-2.18)
or at follow-up (crude OR= 1.50; 99% CI 0.90-2.52, adjusted OR= 1.54; 99% CI 0.86-2.75).
D
Obesity at baseline was associated with GERS (crude OR= 1.86; 99% CI 1.03-3.37, adjusted
TE
OR=1.96; 99% CI 1.05-3.64) but not at follow-up (crude OR=1.43; 99% CI 0.74-2.77, adjusted
at baseline did also overlap with IBS in 56 individuals and at follow-up in 59 cases.
GERS overlapped with IBS in 112 individuals at baseline and in 117 individuals at follow-up.
Suspected and major anxiety were not associated with overlap of FD with GERS or IBS, nor
with overlap of GERS with IBS at baseline (Table 2). Major anxiety was associated with all
overlaps at follow-up (Table 3). Suspected depression or major depression, smoking, use of
13
ACCEPTED MANUSCRIPT
alcohol, H. pylori infection, use of NSAIDs or abnormal weight were not associated with
A total of 33 individuals changed from the FD group at baseline to GERS at follow-up and also
20 individuals changed from the FD group at baseline to IBS at follow-up (11 overlapping with
PT
change to GERS) 17 individuals from the GERS group changed to IBS. Change from FD at
RI
baseline to GERS at follow-up was associated with major anxiety at follow-up (crude
SC
p<0.001). Major anxiety was also associated with change from FD at baseline to IBS at follow-
U
exact p=0.005) and with change from GERS to IBS (crude OR=28.64; 99% CI 2.39-342.88,
AN
adjusted OR= 7.35; 99% CI 2.21-338.23, Fischers exact p=0.002).
M
Both suspected anxiety and major anxiety at baseline were associated with new onset FD
D
(n=48) at follow-up (adjusted OR=3.97; 99% CI 1.20-13.20 and adjusted OR=7.61; 99% CI
TE
1.21-47.73, respectively) but not with new onset GERS (n=93) (adjusted OR=1.67; 99% CI
0.49-5.67 and adjusted OR=1.78; 99% CI 0.35-8.98, respectively). Depression, smoking, high
EP
consumption of alcohol H. pylori infection, use of NSAID/aspirin and obesity at baseline were
There were 14 new onset EPS and 37 new onset PDS and overlap of these occured in 3
individuals. Suspected and major anxiety at baseline were not associated with new onset PDS
(adjusted OR=3.71; 99% CI 0.99-14.03 and adjusted OR=6.48; 99% 0.91-45.75, respectively).
There were no associations of suspected or major anxiety at baseline with new onset EPS
14
ACCEPTED MANUSCRIPT
(adjusted OR=4.15; 99% CI 0.67-25.09 and adjusted OR=4.11; 99% CI 0.23-73.23,
respectively).
Mean anxiety score at baseline for cases where FD disappeared at follow-up was 4.4 and for
those cases still having FD the mean anxiety score at baseline was higher at 6.1 (p=0.021).
PT
Notably, the mean anxiety score at follow-up for those whose FD disappeared was 3.8 versus
RI
those still having FD where the mean anxiety score was significantly higher at 6.9
(p=0.0002).
U SC
AN
M
D
TE
C EP
AC
15
ACCEPTED MANUSCRIPT
DISCUSSION
We have performed a prospective 10 year follow-up study in a well characterized cohort from a
clearly defined general population in Sweden who underwent a baseline endoscopy and
detailed symptom evaluation.(34) To our knowledge, this is the first follow-up of a general
population cohort where FD was defined according to the Rome III definition.(17) We found
PT
that anxiety at baseline was associated with over a seven fold risk of new onset FD symptoms
RI
but not GERS supporting a strong association with FD. Furthermore, a lower anxiety score at
follow-up was associated with the disappearance of FD. In addition, anxiety was associated
SC
with an increased risk of change of FD to GERS or IBS.
U
Our results are in contrast to results from an Australian follow-up study where higher anxiety
AN
score did predict the onset of other functional gastrointestinal disorders but not onset of FD
M
symptoms, although this may be explained by a type II error and none of these subjects had
systematically undergone prior endoscopy.(19) Our results are consistent with earlier
D
FD.(18)
EP
The mechanisms by which anxiety may drive upper gastrointestinal symptoms are likely
C
including delayed gastric emptying(44, 45) as well as reduced proximal gastric relaxation
the relationship between specific symptoms and gastric physiological changes has been
inconsistent.(44-46) Fischler et al. observed that epigastric pain and burning were significantly
associated with both hypersensitivity to gastric distension and the personality trait of
neuroticism which is closely related to anxiety.(48) Geeraerts et al. observed during induced
16
ACCEPTED MANUSCRIPT
anxiety by recalling a neutral or an anxious autobiographical experience that gastric
compliance significantly decreased with higher anxiety levels; intraballoon pressures inducing
discomfort during gastric distention were not altered, but the corresponding volume was
significantly lower. Meal-induced relaxation was inhibited during anxiety and this persisted
for the whole 60-minute measurement.(49) A Dutch study of 149 patients with FD showed that
PT
higher levels of state anxiety predicted lower discomfort and pain thresholds and lower
compliance applying the gastric barostat.(50) Thus the presence of anxiety may predispose to
RI
FD, and when pathophysiological abnormalities are induced for example by infectious
SC
gastroenteritis(51, 52), a vicious cycle may be set up inducing new upper gut symptoms that in
turn worsen anxiety and increase the perception of symptoms leading to the clinical syndrome
recognised as FD.
U
AN
M
We have earlier observed an association of duodenal eosinophilia with PDS and early satiation
in adults from the Kalixanda studies(53) now confirmed worldwide.(54, 55) In the present study
D
we observed an association of anxiety with PDS both at baseline and at follow-up. Anxiety was
TE
not associated with the new onset PDS but the analysis was conservative (applying 99%
confidence limits) and the lack of association may represent a type II error once FD is
EP
response that in turn activates eosinophils, promotes eosinophil degranulation and leads to the
C
Alternatively, anxiety might drive through the stress response immune activation from a TH1
children, an association between anxiety and duodenal eosinophil density has been reported.(59)
17
ACCEPTED MANUSCRIPT
The data support the view anxiety may drive gut to brain as well as brain to gut mechanisms
In this study 187 individuals had persistent GERS pointing to the chronicity of this condition.
We have earlier shown that GERS were more stable (81%), in comparison to the other
PT
gastrointestinal symptom groups in short term.(61) Based on earlier population-based cross-
RI
sectional studies, it is controversial whether an association of anxiety exists with
gastroesophageal reflux disease.(20, 21) Our study of a well characterized population cohort
SC
supports the finding of an association between anxiety and GERS in cross-sectional analyses
only at follow-up. However, importantly there appears not to be a causal link as anxiety did
U
not precede GERS. There are data showing that anxiety heightens sensory perception in
AN
GERS and might reduce the symptomatic response to therapy.(20, 62) Sharma et al. for
M
hyperalgesia.(63)
TE
EP
It is well established that FD, IBS and GERS overlap more than expected by chance but this
overlap remains unexplained.(61, 64) A link to post infectious gastroenteritis has been identified
C
in some cases(51, 52) and a Korean study suggested that anxiety was linked to the overlap
AC
between GERD, dyspepsia and IBS.(65) Our findings at follow-up are consistent with these
results. We speculate the extent of immune activation in the small intestine after exposure to
infectious or food antigens may account for whether patients experience upper or lower
intestinal symptoms; if confined proximally, dyspepsia may be more likely. More distal
intestinal inflammation may induce IBS while while those having extensive proximal and distal
18
ACCEPTED MANUSCRIPT
involvement may be more likely to experience overlapping symptoms.(66, 67) If this is true,
anxiety through the stress response may play a key role in exaggerating the immune process
PT
We did not find an association of NSAIDs with FD. Overall 41 individuals out of 703 (5.8%)
RI
used NSAIDs at baseline, and only 11 of 110 individuals with FD (10%) in this cohort reported
use of NSAIDs at baseline. The low reported use of NSAIDs possibly explains the lack of
SC
association.
U
The present population-based study had a number of strengths. A true general population
AN
based design was applied with a very high participation rate both at baseline and at the ten-
M
year follow-up. Further, the questionnaires used have been found to be robust and
reproducible.(31-33) A possible weakness is that the ASQ was designed before the Rome era
D
but the questionnaire has been updated to meet the Rome II and III criteria and it measures all
TE
of the Rome III criteria aside from the symptom onset (3 vs. 6 months).(31) The ASQ has also
been revalidated after being populated with the additional questions.(32) Performing the
EP
endoscopy only at baseline is a potential weakness but this method has been used earlier in
follow-up studies(14, 68) and ethically we couldnt perform an invasive procedure on the same
C
AC
population twice without a medical indication. Another possible weakness is a slightly lower
response rate to the question on present weight (81.8% of the responders and 82.8% of those
reporting GERS). This might have resulted in under-reporting high BMIs when analyzing the
association of obesity with GERS at follow-up. Another possible weakness is that we do not
have exact data on prescriptions of eradication therapy for H. pylori. However, the local
Swedish guidelines for eradication have remained very restricted (i.e. in peptic ulcer cases
19
ACCEPTED MANUSCRIPT
only). We did not have access to the medical records at the 10-year follow-up nor medicine
codes (ATC codes were available at baseline only) and therefore we could not check
treatment for anxiety and depression at follow-up. PPIs were available only by prescription at
baseline but could be bought over the counter at follow-up and the use of PPIs was markedly
increased at follow-up based on questionnaire responses. This fact might have influenced the
PT
reported EPS and GERS rates but not PDS; PPIs are not efficacious in dysmotility-like
RI
gap in data but this is not possible to avoid in a follow-up study of this kind. Finally, 99%
SC
confidence intervals were chosen to reduce type I errors but it is a conservative approach.
U
In conclusion, anxiety but not depression precedes the onset of new FD but not GERS over a
AN
ten year follow-up suggesting a strong association. Further exploration of the mechanisms
M
linking anxiety with FD may yield new approaches to interventions that reduce or eliminate
symptoms.
D
TE
C EP
AC
20
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
Figure 1.
Figure 2.
PT
Figure 3.
RI
Change of FD, GERS and IBS from baseline to follow-up
U SC
AN
M
D
TE
C EP
AC
21
ACCEPTED MANUSCRIPT
REFERENCES
(Erratum: 2003;21:148).
PT
reflux disease: a systematic review. Gut 2005;54:710-7.
RI
3. Wahlqvist P, Reilly MC, Barkun A. Systematic review: the impact of gastro-
SC
2006;24:259-72.
U
and Functional Dyspepsia (Rome III Criteria) in a Swedish Population-Based
AN
Study. Gastroenterology 2009;137:94100.
M
5. Drossman DA, Richter JE, Talley NJ, Thompson WG, Corazziari E, Whitehead
Gastroenterology 2006;130:1466-79.
8. Zagari RM, Law GR, Fuccio L, et al. Epidemiology of functional dyspepsia and
Gastroenterology 2010;138:1302-11.
reflux symptoms and esophagitis with or without symptoms in the general adult
22
ACCEPTED MANUSCRIPT
Swedish population: A Kalixanda study report. Scand J Gastroenterol
2005;40:275-85.
PT
11. Koloski NA, Talley NJ, Boyce PM. Does psychological distress modulate
RI
functional gastrointestinal symptoms and health care seeking? A prospective,
SC
12. lander T, Svardsudd K, Johansson SE, et al. Psychological illness is commonly
U
during patient consultation: a population-based study. BMC Med 2005;3:8.
AN
13. Koloski NA, Boyce PM, Talley NJ. Is health care seeking for irritable bowel
M
15. Li Y, Nie Y, Sha W, et al. The link between psychosocial factors and functional
Gastroenterology 2009;137:23-6.
23
ACCEPTED MANUSCRIPT
18. Kindt S, Van Oudenhove L, Mispelon L, et al. Longitudinal and cross-sectional
19. Koloski NA, Jones M, Kalantar J, et al. The brain--gut pathway in functional
PT
study. Gut 2012;61:1284-90.
RI
20. Jansson C, Nordenstedt H, Wallander MA, et al. Severe gastro-oesophageal
SC
based study. Aliment Pharmacol Ther 2007;26:683-91.
21. Eslick GD, Talley NJ. Gastroesophageal reflux disease (GERD): risk factors,
U
and impact on quality of life-a population-based study. J Clin Gastroenterol
AN
2009;43:111-7.
M
22. Nandurkar S, Talley NJ, Xia H, et al. Dyspepsia in the community is linked to
smoking and aspirin use but not to Helicobacter pylori infection. Arch Intern
D
Med 1998;158:1427-33.
TE
23. Nocon M, Labenz J, Willich SN. Lifestyle factors and symptoms of gastro-
2006;23:169-74.
24. Fujiwara Y, Kubo M, Kohata Y, et al. Cigarette smoking and its association with
C
AC
25. Walker MM, Aggarwal KR, Shim LS, et al. Duodenal eosinophilia and early
27. Ofman JJ, Maclean CH, Straus WL, et al. Meta-analysis of dyspepsia and
PT
28. Lagergren J, Bergstrom R, Nyren O. No relation between body mass and gastro-
RI
oesophageal reflux symptoms in a Swedish population based study. Gut
2000;47:26-9.
SC
29. El-Serag HB, Graham DY, Satia JA, et al. Obesity is an independent risk factor
2005;100:1243-50.
U
AN
30. Talley NJ, Quan C, Jones MP, et al. Association of upper and lower
M
1993;11:252-62.
EP
32. Aro P, Ronkainen J, Storskrubb T, et al. Validation of the translation and cross-
Hospital Anxiety and Depression Scale and the Complaint Score Questionnaire.
33. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta
PT
negative test for current Hp infection: A report on histology and Helicobacter
RI
pylori detection in the general adult population. Scand J Gastroenterol
2005;40:302-11.
SC
36. Ronkainen J, Aro P, Storskrubb T, et al. Prevalence of Barretts Esophagus in
2005;129:1825-31.
U
AN
37. Vakil N, Van Zanten SJ, Kahrilas P, et al. The Montreal Definition and
M
38. Agrus L, Talley NJ, Svrdsudd K, et al. Identifying dyspepsia and irritable
TE
bowel syndrome: the value of pain or discomfort, and bowel habit descriptors.
39. Brandt LJ, Chey WD, Foxx-Orenstein AE, et al. An Evidence-Based Position
2009;104:S1-35.
40. Ford AC, Bercik P, Morgan DG, et al. Validation of the Rome III criteria for the
2013;145:1262-70 e1.
26
ACCEPTED MANUSCRIPT
41. World Health Organization W. Obesity; preventing and managing the global
1997.
42. Kleinbaum DG. Logistic Regression. A Self-Learning Text. New York, USA:
PT
43. StataCorp. The Intercoold STATA 8 program. In. College Station, Texas, 2003.
RI
44. Stanghellini V, Tosetti C, Paternico A, et al. Risk indicators of delayed gastric
SC
1996;110:1036-42.
U
gastric emptying of solids and liquids in functional dyspepsia. Am J
AN
Gastroenterol 2003;98:783-8.
M
46. Salet GA, Samsom M, Roelofs JM, et al. Responses to gastric distension in
2001;121:526-35.
EP
Gastroenterology 2005;129:1437-44.
27
ACCEPTED MANUSCRIPT
50. Van Oudenhove L, Vandenberghe J, Geeraerts B, et al. Relationship between
Med 2007;69:455-63.
PT
gastroenteritis outbreak. Am J Gastroenterol 2012;107:891-9.
RI
52. Pike BL, Porter CK, Sorrell TJ, et al. Acute gastroenteritis and the risk of
SC
2013;108:1558-63;quiz 1564.
53. Talley NJ, Walker MM, Aro P, et al. Non-ulcer Dyspepsia and Duodenal
U
Eosinophilia: An Adult Endoscopic Population-Based Case-Control Study. Clin
AN
Gastroenterol Hepatol 2007;5:1175-83.
M
56. Zheng PY, Feng BS, Oluwole C, et al. Psychological stress induces eosinophils
9.
57. Liebregts T, Adam B, Bredack C, et al. Small bowel homing T cells are
Am J Gastroenterol 2011;106:1089-98.
28
ACCEPTED MANUSCRIPT
58. Chrousos GP. Stress, chronic inflammation, and emotional and physical well-
2000;106(5 Suppl):S275-91.
59. Schurman JV, Singh M, Singh V, et al. Symptoms and subtypes in pediatric
PT
functioning. J Pediatr Gastroenterol Nutr 2010;51:298-303.
RI
60. Friesen CA, Schurman JV, Colombo JM, et al. Eosinophils and mast cells as
SC
Pharmacol Ther 2013;4:86-96.
U
fluctuate in the short-term perspective? The Kalixanda study. Dig Dis
AN
2008;26:256-63.
M
64. Talley NJ, Dennis EH, Schettler-Duncan VA, et al. Overlapping upper and lower
65. Lee SY, Lee KJ, Kim SJ, et al. Prevalence and risk factors for overlaps between
66. Walker MM, Warwick A, Ung C, et al. The role of eosinophils and mast cells in
PT
symptoms, mental distress and fear of serious illness. Aliment Pharmacol Ther
RI
2006;24:1241-6.
69. Lacy BE, Talley NJ, Locke GR, 3rd, et al. Review article: current treatment
SC
options and management of functional dyspepsia. Aliment Pharmacol Ther
2012;36:3-15.
U
Author names in bold designate shared co-first authorship.
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
Table 1. Association of anxiety and depression at baseline and at follow-up with FD and
GERS at baseline and at follow-up, respectively.
FD FD GERS GERS
at baseline at follow-up at baseline at follow-up
n=110 n=93 n=185 n=203
Suspected anxiety
Crude OR 2.44 1.96 2.00 1.79
PT
99% CI 1.02-5.87 0.71-5.41 0.95-4.20 0.79-4.06
RI
99% CI 0.73-4.87 0.48-4.34 0.77-3.84 0.61-3.86
Major Anxiety
Crude OR 3.67 7.32 1.24 7.83
SC
99% CI 0.97-13.95 2.23-24.10 0.39-3.94 1.90-32.20
U
Suspected depression
AN
Crude OR 0.62 2.14 0.99 1.13
99% CI 0.12-3.11 0.63-7.25 0.36-2.75 0.39-3.23
PT
Adjusted OR 2.78 2.15 2.66
99% CI 0.81-9.54 0.59-7.80 0.97-7.30
RI
Major anxiety
Crude OR 2.35 3.09 1.96
99% CI 0.39-13.96 0.53-18.05 0.45-8.56
SC
Adjusted OR 2.43 3.11 1.97
99% CI 0.40-14.90 0.51-18.85 0.45-8.67
Use of PPI
U
Adjusted OR 28.84 8.21 15.05
99% CI 4.90-169.86 2.22 30.39 1.96-115.84
AN
Male gender
OR 0.43 0.37 0.72
99% CI 0.19-0 .96 0.16-0.88 0.39-1.33
M
Age 60
OR 0.91 0.91 1.00
99% CI 0.40-2.07 0.39-2.14 0.52-1.92
D
PT
99% CI 0.55-6.62 0.66-6.54 0.64-4.74
RI
99% CI 0.35-6.10 0.48 5.43 0.41-3.98
Major anxiety
Crude OR 11.68 5.42 8.36
SC
99% CI 2,55-53.50 1.21-24.21 2.05-34.11
U
Use of PPI
Adjusted OR 10.04 3.34 11.43
AN
99% CI 4.05-24.88 1.43-7.80 5.06-25.80
Male gender
OR 0.40 0.29 0.44
M
PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT
PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT
PT
RI
U SC
AN
M
D
TE
EP
C
AC