Alimentary Pharmacology and Therapeutics

Dyspepsia is strongly associated with major depression and

generalised anxiety disorder - a community study
A. D. P. Mak*, J. C. Y. Wu, Y. Chan, F. K. L. Chan, J. J. Y. Sung & S. Lee

*Department of Psychiatry, Shatin

Hospital, Hong Kong, China.

Institute of Digestive Disease, The

Chinese University of Hong Kong,
Hong Kong, China.

Hong Kong Mood Disorders Centre,

The Chinese University of Hong Kong,
Hong Kong, China.

Correspondence to:
Dr J. C. Y. Wu, Department of
Medicine & Therapeutics, Prince of
Wales Hospital, The Chinese
University of Hong Kong, Shatin,
Hong Kong, China.
E-mail: justinwu@cuhk.edu.hk

Publication data
Submitted 20 February 2012
First decision 23 March 2012
Resubmitted 14 August 2012
Accepted 14 August 2012
EV Pub Online 8 September 2012

SUMMARY
Background
The relationship between dyspepsia and psychiatric comorbidity such as
anxiety and depression is poorly dened. Previous studies have been limited
by lack of standardised diagnostic criteria.
Aim
To examine the prevalence and comorbidity of dyspepsia as dened by
Rome III (6-month duration) with DSM-IV-TR generalised anxiety disorder
(GAD) and major depressive episodes (MDE) in the general population.
Methods
A random population-based telephone survey was done using a questionnaire on symptoms of Rome III Dyspepsia, DSM-IV-TR GAD and MDE
and their chronological relationship.
Results
Of the 2011 respondents 8.0% currently had Rome III Dyspepsia, 3.8%
reported GAD and 12.4% reported MDE respectively. Dyspeptic subjects
had a twofold increased risk of GAD (OR = 2.03, 95% CI: 1.063.89,
P < 0.001) and a threefold increased risk of MDE (OR = 3.56, 95% CI:
2.335.43, P < 0.001). MDE and GAD most often coincided with dyspepsia
in onset. Dyspepsia (OR = 2.48, 95% CI: 1.653.72 P < 0.001), MDE
(OR = 2.39, 95% CI: 1.643.46, P < 0.001) and female sex (OR = 1.65,
95% CI: 1.212.23, P < 0.001) independently predicted frequent medical
consultations. GAD independently predicted high investigation expenditure
(OR = 4.65, 95% CI: 1.1518.70, P = 0.03).
Conclusions
With stringently adopted Rome III and DSM-IV-TR criteria, dyspepsia was
strongly associated and often coincident in onset with generalised anxiety
disorder and major depressive episodes in the community. Excessive healthcare utilisation should alert clinicians to risk of psychiatric comorbidity.
Aliment Pharmacol Ther 2012; 36: 800810

800

2012 Blackwell Publishing Ltd

doi:10.1111/apt.12036

Community study on dyspepsia, depression and anxiety

INTRODUCTION
Dyspepsia is a common gastrointestinal condition that is
characterised by chronic recurrent epigastric symptoms
such as pain, burning and postprandial symptoms. Dyspepsia is associated with signicant functional impairment and burden on healthcare resources.1 Estimates of
community prevalence of dyspepsia have varied widely
from 5% to 40%2 depending on criteria used. Yet,
community prevalence measured using strictly applied
Rome III criteria,3 which are more precise and restrictive
than their predecessors, remains sparse.4
It has been reported that comorbid psychiatric disorders are common in patients with dyspepsia. Generalised
anxiety disorder (GAD) and Major depressive episodes
(MDE), in particular, have been most extensively related
to dyspepsia. These three disorders share several
commonalities. Similar to functional dyspepsia, both
GAD and MDE are typically chronic and relapsing, more
common in female, with onset and exacerbation often
associated with psychosocial stress.5, 6 Moreover, visceral
hypersensitivity and somatisation have been associated
with all three conditions.4, 710
Studies have found increased anxiety and depressive
symptoms on rating scales,11 and poorer psychological
wellbeing12 in organic and functional dyspepsia. However, their relevance to how specic comorbid mental
disorders may contribute to the clinical and societal
burden of dyspeptic symptoms12 remains unclear. As
effective interventions exist for mental disorders,
their treatment may benet the clinical outcome of
dyspepsia.13
To date, most studies on psychiatric correlates in dyspeptic patients came from referral centres and clinics.4
Selection bias may distort any association of psychiatric
morbidity and dyspepsia found in these samples, because
only 2542%14 of dyspepsia sufferers in the community
seek medical care. It is unclear whether psychological
distress may motivate or hinder patients in seeking care
for dyspeptic symptoms. The recent community-based
Kalixanda study15 evaluated the relationship between
anxiety, depression and functional dyspepsia. It found
anxiety, but not depression, to be associated with functional and organic dyspepsia. However, the Rome III
criteria used in this study were post hoc rated from a
Rome II questionnaire with the Rome III 6-month onset
criterion discarded. Meanwhile, HADS, as with most
psychiatric symptom scales, captured symptoms as a
dimensional score, and did not meet the duration
and impairment requirements for diagnosis of GAD
and MDE as mental disorders dened by the American
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Psychiatric Associations Diagnostic and Statistical

Manual of Mental Disorders, Fourth Edition- Text
Revision (DSM-IV-TR)16 (Tables 2 and 3). Using the
Structured Clinical Interview for DSM-IV Axis-I
Disorder (SCID-I),17 we found 38.2% and 16.4% of
Chinese patients in a tertiary gastroenterology clinic with
functional dyspepsia to have an anxiety disorder and
depressive disorder respectively.18
We therefore set out to conduct a community survey
to examine the current community prevalence of stringently dened Rome III Dyspepsia (Table 1), as well as
the 12-month prevalence of GAD and MDE using
DSM-IV-TR. We also evaluated the association between
dyspepsia, GAD and MDE and the chronological
relationship of their onsets. Although rarely examined
in previous studies, the latter may shed light on
whether these illnesses may be causally related to each
other.19

METHODS
A random community-based telephone survey of the
general population in the age range of 1565 years was
conducted from 22nd April to 13th May, 2009. The
study was approved by the research ethics committee of
The Chinese University of Hong Kong. The Hong Kong
Institute of Asia-Pacic Studies of the Chinese University
of Hong Kong, an independent survey research organisation, was commissioned to conduct the survey. Interviewers were university students with 13 years of
part-time experience in telephone interviews. A brieng
session was held by the investigators (AM, JW, YC) for
training of administration of the questionnaire and
explaining skills involved in eliciting symptoms of dyspepsia, GAD and MDE. Respondents were invited to
take part in a telephone survey of digestive problems
and emotional health. Interviews proceeded only after
verbal consent was sought. The interviews were
conducted in Cantonese dialect, the predominant spoken

Table 1 | Rome III symptomatic criteria for functional

dyspepsia3
Must include
One or more of
a Bothersome postprandial fullness
b Early satiation
c Epigastric pain
d Epigastric burning
Criteria fullled for the last 3 months with symptom onset at
least 6 months before diagnosis.
801

A. D. P. Mak et al.
Table 2 | Survey questions pertaining to the DSM IV-TR symptoms of generalised anxiety disorder
Required response for
diagnosis

DSM-IV criteria

Questions in the telephone survey

A. Excessive anxiety and worry, occurring

more days than not for at least 6 months,
about a number of events or activities
(e.g. work or school performance)

A1. Do you think that you are prone to

anxiety or worries?

At least one response

to A1A3 has to be
Yes

A2. Was there a period in the past twelve

months that you were habitually worried?
A3. In the past twelve months, had you
frequently worried about a number of
events or activities (e.g. work or school
performance)?
A4. Did your worries or anxiety last more
than 6 months?
A5. During the period when you were
worried or anxious, did you experience
worries or anxiety for most of the time?
B1. During the period when you were
worried or anxious, how often were you
able to control your worry?
C1. Did you frequently experience the
following symptoms when you were
worried or anxious:
1. Restlessness
2. Being easily fatigued
3. Irritability
4. Muscle tension
5. Difculty falling or staying asleep, or
restless unsatisfying sleep
6. Difculty concentrating or the mind going
blank

Yes

B. The person nds it hard to control the

worry
C. The anxiety and worry are associated
with three (or more) of the following six
symptoms (with at least some symptoms
present for more days than not for the
past 6 months)
1. Restlessness or feeling keyed up or on edge
2. Being easily fatigued
3. Irritability
4. Muscle tension
5. Difculty falling/staying asleep, or restless
unsatisfying sleep
6. Difculty concentrating or the mind going blank
D. Clinically signicant distress or
impairment in social, occupational/other
important areas of functioning

D1. Did the above experiences cause you

signicant distress? AND/OR
D2. Did the above experiences signicantly
impair your daily life (e.g. study/work,
social & family life?

dialect in Hong Kong. On average, an interview took

approximately 20 min to nish.
Hong Kong has a population of 7.0 million. More
than 99% of households have a telephone line.20 Sampling telephone lines should therefore generate a representative sample of households. Phone calls were made
in the evening (18:0022:00 hours) to cover the working
population. Of the 6378 valid home telephone numbers
randomly selected from the Hong Kong Telephone
Directory, 2585 hung up immediately, and 720 had no
interviewee in the suitable age range. Of the 3057
successful contacts, 1062 were rejected. Of the 1062
rejected participants, 931 were turned down by relatives
of the suitable respondent, 82 were turned down by the
802

Yes

Yes
Yes
Occasionally or Very
difcult or Unable
to control
At least three of the
six symptoms have
to be met

Very distressed or
Quite distressed
AND/OR
Very impaired or
Quite impaired

suitable respondent and for the remaining 49, contact

could not be made to the suitable respondent after initial
successful contact (Figure 1). This resulted in a nal
sample of 2011 respondents (937 men, 1074 women; age
distribution in years: 1524 (16.9%), 2534 (20.4%), 35
44 (22.6%), 4554 (26%), 5565 (20.7%). The distribution is highly comparable to that reported by the Census
and Statistics Department20 (Table 4). The participation
rate was therefore 65.4%.

Instrument
A 40-item questionnaire was used covering demographic
information, dyspepsia using the Chinese version of the
Rome III Dyspepsia Module,21 symptoms of GAD and
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Community study on dyspepsia, depression and anxiety

Table 3 | Survey questions pertaining to the DSM IV-TR symptoms of major depressive episode
Required response for
diagnosis

DSM IV criteria

Questions in the telephone survey

FIVE or more of the following in the same

2 week period, representing a change from
previous functioning. At least ONE should
be either (1) or (2)
1. Depressed mood most of the day, nearly
every day (reported or observed)
2. Markedly diminished interest or pleasure
in (almost) all activities most of the day,
nearly every day (reported or observed)
3. Signicant weight loss
4. Insomnia or hypersomnia nearly every
day
5. Psychomotor agitation or retardation
nearly every day (observed + subjective)
6. Fatigue or loss of energy nearly every day
7. Feelings of worthlessness or excessive or
inappropriate guilt nearly every day (not
merely self-reproach or guilt about being
sick)
8. Diminished ability to think or
concentrate, or indecisiveness, nearly
every day (subjective or objective)
9. Recurrent thoughts of death (not just fear
of dying), recurrent suicidal ideation
without a specic plan, or a suicide
attempt or a specic plan for committing
suicide.

A1. In the past year, have you had 2 weeks

or longer when you had depressed mood
most of the day, nearly every day?
A2. In the past year, have you had 2 weeks
or longer when you had markedly reduced
interest or pleasure in almost all activities
most of the time, nearly every day?

Yes in A1 or A2

B. During the period when you had

(depressed mood) (reduced interest or
pleasure), did you often have
a Poor appetite or lost weight
b Slept less than usual
c Observable slowing of speech or actions
d Fatigue or loss of energy
e Difculty to concentrate or make
decisions
f Observable restlessness and agitation
g Increased appetite or body weight
h Slept more than usual
i Felt worthless
j Thoughts of death or suicide
D1. Did the above experiences cause you
signicant distress? AND/OR

a and g counted as 1
item
b and h counted as 1
item
c and f counted as 1
item
Yes in at least FIVE
items in sections A and
B, including at least one
item in A.

D. Clinically signicant distress or

impairment in social, occupational/other
important areas of functioning
D. Clinically signicant distress or
impairment in social, occupational/other
important areas of functioning

D2. Did the above experiences signicantly

impair your daily life (e.g. study/work,
social & family life?

MDE (Tables 2 and 3) based on DSM-IV-TR Criteria

and healthcare utilisation.
The same questionnaire for GAD and MDE diagnoses
had been used in previous series of telephone surveys
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Very distressed or
Quite distressed AND/
OR
Very impaired or Quite
impaired

that found 12-month prevalence of GAD to be 4.1%22

and 8.4% for major depressive episode (MDE).23 Clinical
re-appraisal for diagnoses of MDE and GAD using the
gold standard of psychiatric diagnosis in research, the
803

A. D. P. Mak et al.
Phone calls made with
valid telephone numbers
(n = 6378)
Hung up immediately
(n = 2585)
n = 3793
Interview refused by other
family members (n = 931)

achieve 80% power at a signicance level of 0.05, will

have a minimum detectable odds ratio of 2.60. For
the dyspepsia-only subgroup (n = 161), for logistic
regression of binary response variable of GAD (n = 4,
4% GAD without MDE) on the binary independent variable of MDE (n = 53, 33%), the minimum detectable
odds ratio would be 5.24, to achieve 80% power at a signicance level of 0.05.27

n = 2862
No suitable respondent
(n = 720)
n = 2142
Failure to contact
suitable respondent (n = 49)
n = 2093

Analysis
The current prevalence of dyspepsia and 12-month prevalence of generalised anxiety disorder and major depressive episode were presented as percentages. The study
sample was weighted (Appendix S1) according to age and
sex distribution of the Hong Kong general population
based on the 2008 Census data (Table 4). To identify

Refused by suitable
respondent (n = 82)
Successfully interviewed
respondents n = 2011

Figure 1 | Flow diagram depicting various reasons for

and number of subjects who did not participate in the
study.

Structured Clinical Interview for DSM-IV Axis-I Disorders (SCID-I) yielded good agreement.24, 25 The same
series of telephone survey had assessed prevalence of
Rome III Irritable Bowel Syndrome in Hong Kong to be
5.4%, comparable while conservative, with overseas estimates.26
We also asked how much did you spend over the past
5 years on investigations for your gastrointestinal
problems? to assess investigation expenditure and how
frequently did you see a doctor for your health problems
to assess frequency of consultations. Subjects were also
asked for each disorder their respective ages when the
disorder rst came on in their lives so as to estimate
chronological association of the disorders.

Study power
Power calculation was performed post hoc based on the
sample collected to estimate the size of odds ratio that
can be detected in the whole sample and the dyspepsia
subgroup. With an overall sample (n = 2011), a logistic
regression of binary response variable of dyspepsia
(n = 142, 7% dyspepsia without MDE) on the binary
independent variable of GAD (n = 77, 3.8%), to
804

Table 4 | Sociodemographic characteristics of the

respondents and comparison with Hong Kong Census
data

Gender
Male
Female
Age
1524
2534
3544
4554
5565
Education
Primary or below
Secondary
Pre college
College or above
Work status
Employed
Unemployed
Retired
Student
Homemaker
Marital status
Single
Married/living together
Previously married
Income level
HKD  10 000
HKD 10 00030 000
HKD 30 00060 000
 HKD 60 000

Total sample
(n = 2011), % (n)

% Census
2008*

46.6 (937)
53.4 (1074)

47.2
52.8

16.9 (341)
20.4 (411)
22.5 (454)
24.3 (488)
15.8 (317)

17.0
20.4
22.6
24.2
15.8

11.3 (317)
45.2 (905)
10.3 (206)
33.3 (667)

60.6 (1212)
6.4 (129)
7.0 (141)
12.1 (241)
13.8 (278)

38.4 (769)
59.4 (1190)
2.1 (43)

20.8 (389)
46.6 (869)
22.3 (416)
10.3 (192)

* Data from Hong Kong Census and Statistics Department,

2008.
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Community study on dyspepsia, depression and anxiety

factors independently associated with dyspepsia, Chisquared tests were used in the whole sample to explore
the association of the psychiatric diagnoses and sociodemographic variables with dyspepsia. All diagnostic and
sociodemographic variables were then entered into logistic regression (enter mode, without any elimination) with
dyspepsia. Multivariate analysis did not proceed in the
dyspepsia sub-group due to insufcient sample size.
Logistic regression and ordinal regression were used to
examine factors independently associated with high
investigation expenditure [arbitrarily dened as spending
HKD50000 (equivalent to USD6427) or more on investigations over the previous 5 years], frequent medical consultations (arbitrarily dened as monthly or more
frequent medical consultations) and their respective association with all diagnostic and sociodemographic variables. Statistical analyses were performed using the
Statistical Package for Social Studies, Version 16.0 (SPSS
Inc., Chicago, IL, USA). All P values were two-tailed
with the level of statistical signicance specied at 0.05.

RESULTS
Prevalence and sociodemographic prole
Of the 2011 respondents, 8.0% (n = 161) currently fullled Rome III criteria for dyspepsia. Among these
respondents, 81.3% (n = 131) had postprandial distress
syndrome (PDS), 7.5% (n = 12) had epigastric pain syndrome (EPS), 4.3% (n = 7) had both PDS and EPS,
while 14.9% (n = 24) was unclassied (Table 5). All
these subjects reported dyspeptic symptoms that were
occurring more than once a week with onset at least
6 months ago. Amongst all sociodemographic variables,
only lower levels of education were signicantly associated with dyspepsia. There was also a nonsignicant
trend of high prevalence of dyspepsia in female.
(Table 4).
The 12-month prevalence of GAD was 3.8% (n = 77).
The 12-month prevalence of MDE, at 12.4% (n = 249),
has been reported before.28
Comorbidities of dyspepsia
Dyspepsia was signicantly associated with both GAD
and MDE; 21.3% of respondents with MDE met Rome
III criteria for dyspepsia vs. 6.1% of non-MDE respondents (P < 0.001). 32.9% of dyspeptic respondents vs.
10.6% of nondyspeptic respondents met criteria for
MDE (P < 0.001) (Table 6).
In all 24.3% of GAD respondents had dyspepsia vs.
7.3% of non-GAD respondents (P < 0.001); 11.8% of
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Table 5 | Prevalence of dyspepsia and its subgroups,

GAD and MDE
Prevalence,
% (n)
Dyspepsia (Rome III)
Postprandial distress syndrome
Epigastric pain syndrome
Psychiatric diagnoses
Generalised anxiety disorder
Major depressive episode
Number of psychiatric diagnoses
No psychiatric diagnosis
One psychiatric diagnosis (GAD OR MDE)
Two psychiatric diagnoses (GAD and MDE)

8.0 (161)
81.3 (131)
7.5 (12)
3.8 (77)
12.4 (249)
86.5 (1739)
10.9 (219)
2.6 (53)

dyspepsia respondents also had GAD, vs. 3.1% of nondyspepsia respondents (P < 0.001) (Table 6).
In the multivariate analysis that included all sociodemographic correlates, the adjusted odds ratio of GAD being
associated with dyspepsia was 2.03 (95% CI: 1.063.89,
P = 0.03), while the adjusted OR of MDE being associated
with dyspepsia was 3.56 (95% CI: 2.335.43, P < 0.001).

Comorbidities of PDS and EPS

Generalised anxiety disorder was found in 10.6%
(n = 14) of PDS [vs. 3.4% (n = 63) non-PDS, P < 0.001]
respondents and 16.7% (n = 2) of EPS [vs. 3.8%
(n = 75) non-EPS] respondents. MDE was found in
34.4% (n = 45) of PDS respondents [vs. 10.9% (n = 204)
non-EPS] and 25% (n = 3) of EPS [vs. 12.3% (n = 246)
non-EPS] respondents.
Concomitant MDE and GAD
Having concomitant MDE and GAD was associated with
higher prevalence of comorbid dyspepsia than having
only one or none psychiatric diagnosis; 28.3% (n = 15)
of respondents with both GAD and MDE had dyspepsia
(P < 0.001), vs. 19.2% (n = 42) with a single psychiatric
diagnosis (P < 0.001), and 6.0% (n = 104) in those without psychiatric comorbidity. On multivariate analysis,
having both GAD and MDE was signicantly associated
with higher prevalence of dyspepsia (Adjusted
OR = 6.65, 95% CI: 3.3513.21, P < 0.001) than having
one psychiatric diagnosis (Adjusted OR = 3.62, 95% CI:
2.375.51, P < 0.001) and no psychiatric comorbidity.
Chronological relationship in the onset time of
dyspepsia, GAD and MDE
In all, 28.6% of respondents with dyspepsia reported rst
onset of dyspeptic symptoms before age of 20. Current
805

A. D. P. Mak et al.
Table 6 | Sociodemographic characteristics and psychiatric comorbidity in dyspeptic respondents

Gender
Male
Female
Age
1524
2534
3544
4554
5565
Education
Primary or below
Secondary
Pre college
College or above
Work status
Employed
Unemployed
Retired
Student
Homemaker
Marital status
Single
Married/living together
Previously married
Income level
HKD  10 000
HKD 10 00030 000
HKD 30 00060 000
 HKD 60 000
Psychiatric diagnoses
GAD
No GAD
MDE
Number of psychiatric diagnoses
No psychiatric diagnosis
One psychiatric diagnosis
(GAD OR MDE)
Two psychiatric diagnoses
(GAD and MDE)

Dyspepsia
(n = 161), % (n)

Crude OR

Adjusted OR

6.9 (75)
8.9 (86)

1
1.32 (0.951.83)

0.10

1
1.16 (0.781.80)

7.9 (27)
7.1 (29)
6.8 (31)
9.6 (47)
8.5 (27)

0.90 (0.511.56)
0.79 (0.461.37)
0.78 (0.461.34)
1.12 (0.681.84)
1

0.70
0.40
0.37
0.65

1.04 (0.363.02)
0.90 (0.431.88)
0.78 (0.401.53)
1.18 (0.642.18)
1

10.2 (23)
9.1 (82)
7.2 (15)
5.8 (39)

1.87 (1.093.19)
1.61 (1.092.40)
1.24 (0.672.31)
1

0.02*
0.02*
0.50

1.51 (0.723.12)
1.58 (0.992.53)
1.08 (0.542.15)
1

7.6 (92)
8.5 (11)
5.7 (8)
7.9 (19)
10.1 (28)

0.74 (0.471.15)
0.85 (0.411.76)
0.55 (0.251.24)
0.77 (0.421.41)
1

0.18
0.66
0.15
0.39

0.88 (0.511.53)
0.87 (0.391.95)
0.60 (0.231.60)
0.82 (0.282.39)
1

7.5 (58)
8 (95)
14 (6)

1
1.08 (0.771.52)
1.93 (0.774.83)

0.66
0.16

1
0.95 (0.571.60)
1.11 (0.393.15)

9.5 (30.4)
7.9 (69)
6.5 (27)
6.8 (13)

1.44 (0.752.76)
1.17 (0.832.15)
0.95 (0.481.88)
1

0.28
0.62
0.88

0.89 (0.411.87)
0.84 (0.431.65)
0.75 (0.371.54)
1

0.47
0.62
0.95
0.77
0.47
0.60

0.27
0.27
0.06
0.83

0.89
0.27
0.74
0.31
0.72

0.94

0.85
0.85
0.88
0.73
0.62
0.44

24.3 (19)
7.3 (142)
21.3 (53)
6.1 (108)

4.10 (2.377.08)

<0.001*

2.03 (1.063.89)

0.03*

4.11 (2.865.89)

<0.001*

3.56 (2.335.43)

<0.001*

6.0 (104)
26.1 (42)

1
3.72 (2.525.49)

<0.001*

1
3.62 (2.375.51)

<0.001*

<0.001*

9.3 (15)

6.03 (3.2111.33)

<0.001*

6.65 (3.3513.21)

<0.001*

* P < 0.05.

comorbidity with GAD was not signicantly associated

with early onset of dyspeptic symptoms (33.3% dyspeptic
respondents with comorbid GAD vs. 31.3% without comorbid GAD, P = 1.0), nor was MDE comorbidity
(31.4% MDE-comorbid vs. 31.9% non-MDE-comorbid
dyspeptic respondents, P = 1.0).
For respondents with dyspepsia and MDE, 44.5%
reported concomitant onset of the two disorders, 31.5%
reported earlier onset of depression than dyspepsia, while
24% reported onset of dyspepsia before depression. For
806

GAD-dyspepsia comorbid respondents, up to half (54%)

of the respondents reported concomitant onset of the
disorders, while 26.7% reported having dyspepsia rst,
19.3% having GAD rst. The vast majority (89.4%) of
those with both GAD and MDE reported concomitant
onset of GAD and MDE, while 7.8% and 2.8% of those
had GAD and MDE rst respectively.
To further examine the effect of order of onset of dyspepsia and mental disorders on occurrence of dyspepsia,
prevalence of dyspepsia was compared between subjects
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Community study on dyspepsia, depression and anxiety

with onset of mental disorders (GAD or MDE) preceding dyspepsia (To allow comparison, this group also
comprised those with only mental disorders now, based
on the assumption of future onset of dyspepsia) and
those where onset of mental disorders did not precede
dyspepsia. Dyspepsia was not more common in those
with mental disorders preceding dyspepsia (7.7%) compared with those where mental disorders did not precede
dyspepsia (8.0%, P = 0.87). On multivariate analysis that
included all other sociodemographic factors, the association of precedent-onset mental disorders with dyspepsia
remained insignicant (Adjusted OR = 0.83 95% CI:
0.471.45, P = 0.51).

Frequent medical consultations

Frequent medical consultations were, on univariate analysis, less common amongst male (M 10.9% vs. F 17.3%,
P < 0.001) and students. It was more common amongst
homemakers (Employed 14.9%, unemployed 14.1%,
retired 17.1%, students 5%, homemakers 18.8%; P = 0.04).
Respondents with dyspepsia made signicantly more
frequent medical consultations than those with no dyspepsia (31.7% vs. 12.8%, P < 0.001). GAD and MDE
were also associated with more frequent medical consultations (36.8% GAD vs. 13.5% non-GAD, P < 0.001;
30.4% MDE vs. 12.1% non-MDE, P < 0.001). Using
multivariate analysis, dyspepsia (Adjusted OR = 2.48,
95% CI: 1.653.72 P < 0.001), MDE (Adjusted
OR = 2.39, 95% CI: 1.643.46, P < 0.001), female gender
(Adjusted OR = 1.65, 95% CI: 1.212.23, P < 0.001)
were all found to be independent predictors of frequent
medical consultations.
High investigation expenditure
In all, 14% of dyspeptic respondents vs. 9.4% of nondyspeptic respondents spent moderately [dened as more
than HKD10000 (equivalent to USD 1289) on medical
investigations over the past 5 years]. 3.2% dyspeptic vs.
1.3% nondyspeptic respondents had high investigation
expenditure [dened as spending more than hkd 50 000
(equivalent to USD 6449) on medical investigations over
the past 5 years]. There was a statistically nonsignicant
trend of dyspepsia being associated with high investigation expenditure (P = 0.06). GAD (5.6% vs. 1.3%,
P = 0.003), but not MDE (2.5% vs. 1.3%, P = 0.16),
was associated with high investigation expenditure on
univariate analysis. GAD was the only signicant
predictor of high investigation expenditure on multivariate analysis (Adjusted OR = 4.65, 95% CI: 1.1518.70,
P = 0.03).
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

DISCUSSION
We set out to determine the current prevalence of
dyspepsia with stringent Rome III symptom criteria in a
random, representative Chinese community sample.
Participation rate was comparable to previous telephone
health surveys.26 We observed that dyspepsia was
strongly associated and often had coincident onset with
GAD and MDE in the community.
Direct comparisons for community prevalence of
Rome III dyspepsia are sparse. Our conservative gure
was, possible cross-national variation aside, consistent
with the Rome III restrictiveness.3 Omission of the
6-month onset requirement from Rome III criteria may
explain the high prevalence of 20% in the Kalixanda
study.15 We found dyspepsia slightly more common in
females, as in local and overseas data.14, 15 Our GAD
prevalence and sociodemographic correlates were consistent with extant community data, which were also
similar to Western ndings and a previous identically
designed survey in Hong Kong.26 The 12.4% MDE
prevalence, reported previously, was higher than the
2007 gure of 8.5%, probably related to the 2008 global
nancial crisis.28
We found signicant and independent association of
dyspepsia with GAD and MDE as mental disorders. This
is consistent with the reported association of anxiety
symptoms with dyspepsia.4, 15 The link with depression
is more controversial. While we found a strong association between MDE and dyspepsia, even higher than that
for GAD, the Kalixanda study found anxiety but not
depression to be associated with dyspepsia.15 This may
be due to the inclusion of dyspepsia of shorter duration
in the Kalixanda study, which may have a weaker association with depression than anxiety. In addition, the
HADS depression subscale comprises only symptom
scores over 1 weeks duration and had no requirement
for functional impairment, and therefore it was substantially less accurate than DSM-IV-TR MDE, which
required 2 weeks of depressive symptoms causing signicant functional impairment or marked distress (Table 3).
Nonetheless, signicant association between dyspepsia
and depressive symptoms was noted in a previous local
community survey using HADS and Rome II,14 as well
as a meta-analysis.4 Future surveys on psychiatric comorbidity using standard diagnostic assessment will help
clarify the relationship between dyspepsia and depression.
GAD and MDE were both more common in PDS,
while only GAD was signicantly more common in EPS.
In view of the small sample size, between-group
807

A. D. P. Mak et al.
comparison would not be meaningful and as such should
be examined in larger community samples.
Particularly salient was the nding that GAD and MDE,
as mental disorders each having distinct biological and
environmental aetiologies, were associated with dyspepsia.
While cross-sectional studies map chronology poorly, our
preliminary ndings were that most of the comorbid subjects had coincident onset of mental disorders and dyspepsia, and that preceding mental disorders were not a
signicant risk factor for dyspepsia. Coincident-onset thus
appears to be the predominant mode of comorbidity. This
disputes the assertion of distress from dyspepsia itself
totally explaining its association with depressive and anxiety features. It is also against the view of mental disorders
causing dyspepsia. Rather, the nding is consistent with
neurophysiological ndings that autonomic nervous system and hypothalamo-pituitary-adrenal axis derangements typical of anxiety disorders and depressive
disorders29 may alter gastrointestinal function.30 It also
concurs with evidence that onset of dyspeptic symptoms
behaves similarly as anxiety and depressive disorders as
systemic reactions to major stressful life events.31 This
strong association between dyspepsia, GAD and MDE
implies that mental disorders should be routinely screened
during the assessment of patients with dyspepsia.
The stepwise increase in risk of dyspepsia with
increased number of psychiatric diagnoses mirrors
evidence from the psychosomatic literature that supports
a dose-response link between somatic symptom load and
psychopathology.32 Studies with larger samples and
detailed severity measures should conrm whether
dyspeptic symptoms could predict level of psychiatric
morbidity in a proportionate manner. Methodologically,
this stresses the value of including both anxiety and
depressive disorders when studying psychiatric comorbidity of medical disorders.
Psychiatric comorbidity seemed to have a complex
impact on healthcare use. In this study, dyspepsia and
MDE independently predicted frequent medical consultations while GAD predicted high investigation expenditure. The extent of somatic symptoms is unlikely to
explain the difference, as patients with both GAD and
MDE are associated with frequent physical complaints
that would drive healthcare use.32 Variances in doctorpatient interaction and cognitive factors may be other
explanations. In particular, more than 70% of GAD
sufferers have excessive health anxiety,33 which may
drive medical decisions into ordering more investigations. For MDE, medical consultations may be driven by
the distinct distress and impairment in depression, and
808

increased mental health literacy of patients regarding

depression as an illness.
Our study has several limitations. First, endoscopy was
not done. It is uncertain how strictly applied Rome III
dyspepsia symptom criteria would predict lack of organicity. However, the subgroup of subjects with peptic
ulcer disease and gastric ulcer is likely to be a small
minority. Our previous endoscopy study in referred
patients with dyspepsia found organicity in less than 10%
of patients.34 The Kalixanda study found no organicity in
77% of uninvestigated dyspepsia cases,15 while a metaanalysis found positive endoscopy in 20% of dyspeptics,
only 6% in Rome-dened dyspeptics.35 Second, we did a
brief telephone survey instead of a detailed face-to-face
one. However, by avoiding face-to-face contact, telephone
surveys may avoid psychiatric stigma, and facilitate disclosure of sensitive information.36 This is critical for the
Chinese for whom disclosing psychological distress may
be culturally inhibited.37 Third, other potentially important factors associated with dyspepsia were omitted owing
to the need for brevity in a telephone survey, such as
medication use. Fourth, recall bias and cross-sectional
design limited the study of chronology. In particular, our
analysis on order of onset and prevalence of dyspepsia
had a clear limitation in assuming future onset of
dyspepsia for those with only mental disorders currently, but such comorbidity was not inevitable. The
order of onset thus designated was limited by this
assumption and any implications on causality need to
be studied more rigorously in future prospective community studies. Fifth, the study was underpowered for
multivariate analysis for variables associated with psychiatric comorbidity and healthcare utilisation in the
dyspepsia subgroup. Future population-based studies
using larger samples may endeavour to explore this
area. Lastly, we assessed current prevalence of dyspepsia
and 12-month prevalence of GAD and MDE so as to
render the ndings comparable to other community
surveys. It should be noted that this approach may
somewhat inate the prevalence gures for the mental
disorders relative to that of dyspepsia.
In conclusion, we found Rome III dyspepsia to be
common and strongly associated with MDE and GAD as
mental disorders, with substantial impact on healthcare
use in the community. The implications are that one,
mental disorders should be routinely screened in assessing dyspepsia, especially those exhibiting high levels of
healthcare use. Two, clinicians treating dyspepsia with
comorbid GAD or MDE should regard help-seeking
behaviour as a salient part of illness experience. Instead
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

Community study on dyspepsia, depression and anxiety

of routinely discouraging unnecessary consultations or
investigations, holistic and effective approaches in treating these complexly ill individuals may do their health
better while reducing unnecessary expenses. Collaborative psychosomatic care involving gastroenterologists and
mental health professionals may address this. Lastly, prospective research is needed on courses of dyspepsia and
comorbid mental disorders, as well as translational
research in such light.

ACKNOWLEDGEMENTS
Declaration of personal interests: Dr Arthur Mak is supported by an educational grant of Pzer Pharmaceutical.
Dr Justin Wu is supported by research funds of Department of Medicine & Therapeutics and educational grant
of Pzer Pharmaceutical. Dr Sing Lee received educa-

tional grants on mental health education from Pzer

Pharmaceuticals, GlaxoSmithkline, Wyeth Pharmaceuticals and Johnson & Johnson Pharmaceuticals. Declaration of funding interests: None.

SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Appendix S1. Comparison of study sample composition and 2008 Hong Kong Census population data, and
weighting factor applied.
Please note: Wiley-Blackwell are not responsible for
the content or functionality of any supporting materials
supplied by the authors. Any queries (other than missing
material) should be directed to the corresponding author
for the article.

REFERENCES
1. Agreus L, Borgquist L. The cost of
gastro-oesophageal reux disease,
dyspepsia and peptic ulcer disease in
Sweden. Pharmacoeconomics 2002; 20:
34755.
2. El-Serag HB, Talley NJ. Systemic
review: the prevalence and clinical
course of functional dyspepsia.
Aliment Pharmacol Ther 2004; 19:
64354.
3. Tack J, Talley NJ, Camilleri M, et al.
Functional gastroduodenal disorders.
Gastroenterology 2006; 130: 146679.
4. Henningsen P, Zimmermann T, Sattel
H. Medically unexplained physical
symptoms, anxiety, and depression: a
meta-analytic review. Psychosom Med
2003; 65: 52833.
5. Hasin DS, Goodwin RD, Stinson FS,
Grant BF. Epidemiology of major
depressive disorder: results from the
National Epidemiologic Survey on
Alcoholism and Related Conditions.
Arch Gen Psychiatry 2005; 62: 1097
106.
6. Carter RM, Wittchen HU, Pster H,
Kessler RC. One-year prevalence of
subthreshold and threshold DSM-IV
generalized anxiety disorder in a
nationally representative sample.
Depress Anxiety 2001; 13: 7888.
7. Van Oudenhove L, Vandenberghe J,
Geeraerts B, et al. Relationship between
anxiety and gastric sensorimotor
function in functional dyspepsia.
Psychosom Med 2007; 69: 45563.
8. Lehoux CP, Abbott FV. Pain, sensory
function, and neurogenic inammatory
response in young women with low
Aliment Pharmacol Ther 2012; 36: 800-810
2012 Blackwell Publishing Ltd

9.

10.

11.

12.

13.

14.

mood. J Psychosom Res 2011; 70:

2419.
de Medeiros MT, Carvalho AF, De
Oliveira Lima JW, Dos Santos AA, de
Oliveira RB, Nobre ESMA. Impact of
depressive symptoms on visceral
sensitivity among patients with different
subtypes of irritable bowel syndrome. J
Nerv Ment Dis 2008; 196: 7114.
Blomhoff S, Spetalen S, Jacobsen MB,
Malt UF. Phobic anxiety changes the
function of brain-gut axis in irritable
bowel syndrome. Psychosom Med 2001;
63: 95965.
Haug TT, Svebak S, Wilhelmsen I,
Berstad A, Ursin H. Psychological
factors and somatic symptoms in
functional dyspepsia. A comparison
with duodenal ulcer and healthy
controls. J Psychosom Res 1994; 38:
28191.
Koloski NA, Talley NJ, Boyce PM.
Does psychological distress modulate
functional gastrointestinal symptoms
and health care seeking? A prospective,
community Cohort study. Am J
Gastroenterol 2003; 98: 78997.
Pajala M, Heikkinen M, Hintikka J. A
prospective 1-year follow-up study in
patients with functional or organic
dyspepsia: changes in gastrointestinal
symptoms, mental distress and fear of
serious illness. Aliment Pharmacol Ther
2006; 24: 12416.
Hu WH, Wong WM, Lam CL, et al.
Anxiety but not depression determines
health care-seeking behaviour in
Chinese patients with dyspepsia and
irritable bowel syndrome: a population-

15.

16.

17.

18.

19.

20.

based study. Aliment Pharmacol Ther

2002; 16: 20818.
Aro P, Talley NJ, Ronkainen J, et al.
Anxiety is associated with
uninvestigated and functional dyspepsia
(Rome III criteria) in a Swedish
population-based study.
Gastroenterology 2009; 137: 94100.
American Psychiatric Association.
Diagnostic and Statistical Manual for
Mental Disorders, 4th edn. Text
Revision. Washington, DC: American
Psychiatric Association, 2000.
So EK, Kam I, Leung CM, Chung D,
Liu Z, Fong S. The Chinese-bilingual
SCID-I/P Project: stage 1 - reliability
for mood disorders and schizophrenia.
Hong Kong J Psychiatry 2003; 13:
718.
Tse AW, Lai LH, Lee CC, et al.
Validation of self-administrated
questionnaire for psychiatric disorders
in patients with functional dyspepsia. J
Neurogastroenterol Motil 2010; 16: 52
60.
Eaton WW, Kramer M, Anthony JC,
Drman A, Shapiro S, Locke BZ.
Conceptual and methodological
problems in estimation of the incidence
of mental disorders from eld survey
data. In: Copper BH, Helgason T, eds.
Epidemiology and the Prevention of
Survey Data. London: Routledge, 1989;
10827.
Census & Statistics Department tGotH.
Hong Kong Statistics. Hong Kong 2009.
Available at: http://www.statistics.gov.
hk/pub/B10100032009AN09B0700.pdf.
Accessed December 1, 2011.
809

A. D. P. Mak et al.
21. Functional Dyspepsia Module. Tampa,
Florida: Rome Foundation, Inc.
Available at: http://www.romecriteria.
org/pdfs/DyspepMode.pdf. Accessed
December 1, 2011.
22. Lee S, Tsang A, Chui H, Kwok K, Cheung
E. A community epidemiological survey
of generalized anxiety disorder in Hong
Kong. Community Ment Health J 2007;
43: 30519.
23. Lee S, Tsang A, Kwok K. Twelvemonth prevalence, correlates, and
treatment preference of adults with
DSM-IV major depressive episode in
Hong Kong. J Affect Disord 2007; 98:
12936.
24. Lee S, Tsang A, Lau L, Mak A, Ng KL,
Chan DM. Concordance between
telephone survey classication and faceto-face structured clinical interview in
the diagnosis of generalized anxiety
disorder in Hong Kong. J Anxiety
Disord 2008; 22: 140311.
25. Lee S, Tsang A, Mak A, Lee A, Lau L,
Ng KL. Concordance between
telephone survey classication and
face-to-face interview diagnosis of
one-year major depressive episode in
Hong Kong. J Affect Disord 2010; 126:
15560.
26. Lee S, Wu J, Ma YL, Tsang A, Guo WJ,
Sung J. Irritable bowel syndrome is

810

27.

28.

29.

30.

31.

32.

33.

strongly associated with generalized

anxiety disorder: a community study.
Aliment Pharmacol Ther 2009; 30:
64351.
Hsieh FY, Block DA, Larsen MD. A
simple method of sample size
calculation for linear and logistic
regression. Stat Med 1998; 17: 162334.
Lee S, Guo WJ, Tsang A, et al.
Evidence for the 2008 economic crisis
exacerbating depression in Hong Kong.
J Affect Disord 2010; 2: 12533.
Holsboer FK, Kunzel HE. Clinical
neuroendocrinology. In: Charney DS,
Nestler EJ, eds. Neurobiology of Mental
Illness. New York: Oxford University
Press, 2004; 15570.
Tache Y, Bonaz B. Corticotropinreleasing factor receptors and stressrelated alterations of gut motor
function. J Clin Invest 2007; 117: 3340.
Naliboff BD, Mayer M, Fass R, et al.
The effect of life stress on symptoms of
heartburn. Psychosom Med 2004; 66:
42634.
Kroenke K, Jackson JL, Chamberlin J.
Depressive and anxiety disorders in
patients presenting with physical
complaints: clinical predictors and
outcome. Am J Med 1997; 103: 33947.
Lee S, Ma YL, Tsang A. A community
study of generalized anxiety disorder

34.

35.

36.

37.

with vs. without health anxiety in Hong

Kong. J Anxiety Disord 2011; 25: 37680.
Sung JJ, Lao WC, Lai MS, et al.
Incidence of gastroesophageal
malignancy in patients with dyspepsia
in Hong Kong: implications for
screening strategies. Gastrointest Endosc
2001; 54: 4548.
Ford AC, Marwaha A, Lim A,
Moayyedi P. What is the prevalence of
clinically signicant endoscopic ndings
in subjects with dyspepsia? Systematic
review and meta-analysis. Clin
Gastroenterol Hepatol 2010; 8: 8307.
McQueen DV. Comparison of results of
personal interview and telephone
surveys of behavior related to risk of
AIDS: advantages of telephone
techniques. In: Fowler F, ed. Health
Survey Research Methods: Conference
Proceedings. Washington, DC:
Government Printing Ofce, 1989;
24752.
Ho DYF, Fu W, Ng SM. Guilt, shame
and embarrassment: revelations of
face and self. Culture Psychology 2004;
10: 6484.

Aliment Pharmacol Ther 2012; 36: 800-810

2012 Blackwell Publishing Ltd