Download a population-based cohort study

Published by Oxford University Press on behalf of the International Epidemiological Association
ß The Author 2013; all rights reserved.
International Journal of Epidemiology 2013;42:1873–1881
doi:10.1093/ije/dyt213
Increased incidence of herpes zoster in adult
patients with peptic ulcer disease: a
population-based cohort study
Jen-Yin Chen,1,2,3 Tain-Junn Cheng,4,5,6,7,8 Chia-Yu Chang,4 Kuo-Mao Lan,1 Shih-Feng Weng,9
Ming-Jen Sheu,10 Su-Feng Tseng3 and Miao-Lin Hu2,11*
1
Department of Anesthesiology, Chi Mei Medical Center, Tainan, Taiwan, 2Department of Food Science and Applied Biotechnology,
National Chung Hsing University, Taichung, Taiwan, 3Department of the Senior Citizen Service Management, Chia Nan University
of Pharmacy and Science, Tainan, Taiwan, 4Department of Neurology, Chi Mei Medical Center, Tainan, Taiwan, 5Department of
Occupational Medicine, Chi Mei Medical Center, Tainan, Taiwan, 6Department of Management in Medical Records and
Information, Chi Mei Medical Center, Tainan, Taiwan, 7Department of Occupational Safety / Institute of Industrial Safety and
Disease Prevention, College of Sustainable Environment, Chia Nan University of Pharmacy and Science, Tainan, Taiwan,
8
Department of Occupational and Environmental Medicine, National Cheng Kung University Hospital, Tainan, Taiwan,
9
Department of Medical Research, Chi Mei Medical Center, Tainan, Taiwan, 10Department of Gastroenterology, Chi Mei Medical
Center, Tainan, Taiwan and 11Department of Health and Nutrition Biotechnology, Asia University, Taichung, Taiwan
*Corresponding author. Department of Food Science and Applied Biotechnology, National Chung-Hsing University, 250 Kuo-Kuang
Road, Taichung, Taiwan 40227. E-mail: [email protected]; [email protected]
Accepted
16 September 2013
Background Peptic ulcer disease (PUD) has been linked to immunological dysfunctions and malnutrition—both are predictors of herpes zoster
(HZ). We hypothesized that PUD patients might have a greater
risk for developing HZ.
Methods
Using the longitudinal National Health Insurance Research
Database, we conducted a retrospective population-based cohort
study to evaluate the incidence of HZ in adult PUD patients and
controls matched by age and gender. Kaplan-Meier analysis and
Cox regression were conducted to compare differences in the development of HZ. The effects of comorbidities on the risk of HZ and
the associations between different risk factors of PUD and HZ were
assessed by subgroup analyses.
Results
We identified 41 229 adults PUD patients and 41 229 controls. Over
an 8-year follow-up, the cumulative incidence of HZ in PUD patients (11.76/1000 person-years) was significantly higher than controls (6.56/1000 person-years) (P < 0.001, by log-rank test). After
adjusting for potential confounders, PUD was an independent predictor for HZ (adjusted hazard ratio: 1.77, 95% confidence interval
1.64–1.91, P < 0.001). Major risk factors of PUD were non-selective
nonsteroidal anti-inflammatory drugs (nsNSAID) usage and
Helicobacter pylori infection. Subgroup analyses revealed that PUD
is an independent predictor of HZ after excluding the potential
confounding effects of the comorbidities; PUD associated with the
combination of nsNSAID usage and Helicobacter pylori infection leads
to greatest risk of HZ.
Conclusions Adults with PUD are at increased risk of HZ independently compared with the general population. Further studies are required to
clarify the nature of the associations between HZ and PUD.
1873
1874
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Keywords
Retrospective cohort study, peptic ulcer disease, herpes zoster,
Helicobacter pylori, non-selective nonsteroidal anti-inflammatory
drugs, National Health Insurance Research Database
Introduction
Herpes zoster (HZ) is a common disease. HZ is caused
by a reactivation of latent varicella zoster virus and is
characterized by painful rashes. Herpetic pain develops in the acute phase and usually subsides
within 2–4 weeks spontaneously. However, approximately 20–25% of HZ patients experience postherpetic
neuralgia or other nonpain complications. Any complication substantially increases the cost of HZ-related
care. The medical expense of this common disease
leads to a substantial burden on the health care
system. Diminishing cell-mediated immunity is a
well-known risk factor of HZ. Approximately 90% of
HZ cases occur among immunocompetent patients.1–5
An HZ vaccine for adults has been available since
2006.6,7 Identifying potentially unfavourable predictors of HZ may help physicians to identify highrisk patients and administer HZ vaccinations to
reduce the incidence of HZ.
Various diseases linked to impaired immunity such
as rheumatic diseases and cancers have been reported
to be associated with an increased risk of HZ.8,9 Peptic
ulcer disease (PUD) is a gastrointestinal disorder that
is common in HZ patients.10 The predominant risk
factors of PUD are Helicobacter pylori (H. pylori) infection and the use of non-selective nonsteroidal antiinflammatory drugs (nsNSAIDs), accounting for 48
and 24% of cases, respectively, in the USA.11 In addition to nsNSAIDs, other ulcerogenic medications
including aspirin, steroids, chemotherapeutics and
cyclooxygenase-2 inhibitors12 are considered risk factors of PUD. H. pylori infection has been linked to
extradigestive manifestations including malnutrition
and immunological dysregulation.13–16 The usage of
nsNSAIDs, aspirin, steroids and chemotherapeutics
has been demonstrated to inhibit cell-mediated immunity.17,18 PUD leads to chronic protein/blood
loss19 and micronutrients deficiencies,13,16,20 which
consequently impair immunity.21,22 Taken together,
PUD patients commonly are found to have impaired
immunity and nutritional deficiencies. We therefore
hypothesized that PUD patients might have a greater
risk for developing HZ. However, evidence regarding
the association between PUD and the incidence of HZ
is limited and worthy of being explored.
Methods
Data sources
We conducted a retrospective cohort study using the
longitudinal National Health Insurance Research
Database (LNHIRD) containing one million random
subjects (about 4% of the enrollees) from 1996–2009
computed by the Taiwan National Health Insurance
(NHI) programme. There were no statistical differences in age, gender or income between the sample
group and all enrollees. The NHI programme is a nationwide healthcare system in Taiwan, established in
1995. The coverage rate in 2002 was approximately
96% of Taiwan’s entire population.4,9 The NHI electronic data provided encrypted patient identification
numbers, gender, date of birth, the ICD-9-CM
(International Classification of Diseases, Ninth revision, Clinical Modification), prescription drugs dispensed, medical cost, medical care facilities and
specialties. All NHI datasets can be interlinked with
each individual personal identification number.
Our study was exempted from full review by the
Institutional Ethics Committee of Chi Mei Medical
Center because we used de-identified secondary data
released to the public for research purposes.
Study population
In the study cohort, we identified adult patients (aged
18 years and older) with PUD by the following criteria: a primary diagnosis with ICD-9-CM code
530-534 concurrent with a claim of gastrointestinal
endoscopy (a medical procedure covered by Taiwan’s
NHI programme23) and prescriptions for PUD. Based
on the reimbursement policy of Taiwan’s NHI, all patients are strictly required to have documented
H. pylori infection or endoscopy-confirmed ulcers
before receiving H. pylori eradication therapy or antiulcer agents (proton pump inhibitors or histamine2
blockers). Standardized diagnostic methods for detecting H. pylori infection include urea breath test,
culture and rapid urease test. Testing and treatments
for PUD did not change over time from 2000 to 2008.
H. pylori eradication therapy is defined as a triple or
quadruple regimen consisting of the following: a
proton pump inhibitor or histamine2 blocker, plus clarithromycin or tetracycline, plus amoxicillin or metronidazole, with or without bismuth. Furthermore,
patients with a confirmed diagnosis of H. pylori infection are only reimbursed for 2 weeks of antibiotic
therapy.23,24
We excluded patients who had diagnostic codes of
human immunodeficiency virus infection (ICD-9-CM
042, 043, 044), organ transplants (ICD-9-CM 336,
375, 3751, 116, 1160, 1164, 1169, 505, 5059, 528,
5280, 5283, 556, 5569, 335, 3350–3352), chronic obstructive pulmonary diseases (ICD-9-CM 491, 492,
496), rheumatic diseases (ICD-9-CM 710, 714, 725)
PEPTIC ULCER DISEASE INCREASES ZOSTER INCIDENCE
and autoimmune diseases (ICD-9-CM 556.9, 555.9,
250.01, 579, 358, 696.1, 696.0, 710.1, 340, 710.2 and
245.2) from the study cohorts and control cohorts because these patients may have been treated with steroids and immunosuppressive medications or suffered
from immunosuppressive conditions which are potential confounders of HZ.9,25
Patients were identified by an automatched search12
in either inpatient or outpatient service claims from
January 2002 to December 2008. Controls were selected using pair matching by age, gender and year
of cohort entry in a ratio of 1:1 through random computerized selection from the remaining subjects in
the database who had no diagnostic codes with
PUD, and did not take ulcerogenic medications to exclude asymptomatic PUD.26 We collected information
from all study subjects including demographic characteristics, treatment modalities, risk factors and
comorbidities.
Incident HZ cases were identified by an automatch
search for codes for HZ and HZ complications (ICD-9
codes 053.0–053.9) present in either an inpatient or
outpatient service claim and with treatment modalities for HZ.4 Inpatients and outpatients with primary
diagnoses of HZ for the first time between 1 January
2002 and 31 December 2008, were included. We
excluded patients with a diagnosis of HZ between 1
January 1996 and 31 December 2001. From the date
of cohort entry, all patients were followed until developing HZ or until the end of 2009, whichever was
earlier.9 In this study, we used administrative data
alone to define cases of HZ for two reasons. First,
accuracy of the claims database is routinely monitored
by the Bureau of Taiwan’s NHI which performs strict
quarterly expert reviews on random samples of every
50–100 ambulatory and inpatient claims for the purpose of reimbursement.1,27,28 Second, a populationbased study in Taiwan reveals that no specific therapy
is used for the management of HZ in Taiwan.
Common medications used for the treatment of HZ
include antiviral agents (23.6%), NSAIDs (61.1%),
acetaminophen (49.3%), systemic corticosteroids
(14.7%), tricyclic antidepressants (8.9%), anticonvulsants (6.6%) and opiates (3.9%).4
Based on the risk factors of PUD, our study PUD
patients were divided into four subgroups for further
stratified analyses including: (i) the combination of
nsNSAID usage and H. pylori infection; (ii) nsNSAID
usage only; (iii) H. pylori infection only; and (iv)
Others, where cyclooxygenase-2 inhibitors, aspirin,
nonaspirin anticoagulants or other ulcerogenic medications were used.11 H. pylori-associated PUD was
identified by the eradication regimens of H. pylori
during or after the index PUD.12,23
Potential confounders
Comorbidities including diabetes mellitus, cancers
and hypertension25 were selected and identified
in the 12 months before the index ambulatory care
1875
visits, from inpatient and outpatient diagnoses in
both cohorts because these comobidities are potential
confounders for HZ. PUD, diabetes mellitus and cancers were not included in the Charlson comorbidity
index29 (CCI) scores which included myocardial infarction, congestive heart failure, peripheral vascular
disease, cerebrovascular disease, dementia, liver disease, hemiplegia, quadriplegia and renal diseases.
We assessed the rate of CCI scores greater than 1 in
PUD patients by the dichotomized outcome of the
Charlson scores.
Outcome measures
The incidence and hazard ratios (HRs) of HZ over
the follow-up period were computed. We performed
prespecified subgroup analysis on comorbidities because comorbidities are considered confounders.30
Furthermore, we performed a stratified analysis to
assess independent predictors of HZ based on different risk factors of PUD. We also assessed the annual
number of HZ patients and the annual incidence for
HZ in all adult PUD patients during the follow-up
period.
Statistical analysis
Data processing and statistical analysis were performed using SAS statistical software (Version 9.2;
SAS Institute, Cary, NC, USA). Chi-square tests were
used to test the differences in categorical data. The
differences of the CCI scores and months of followup were considered as continuous variables and were
tested by Student’s t-test. Kaplan–Meier analysis was
used to calculate the cumulative incidence of HZ between the two groups and the log-rank test was used
to test the difference between PUD patients and the
controls. The Cox regression model was performed to
evaluate the effects of comorbidities to compare the
HRs potential. The Cox hazard regressions among different risk factors of PUD were performed to evaluate
the effects on HZ in PUD patients. A P value of <0.05
was considered statistically significant.
Results
We identified 41 229 PUD adults and 41 229 matched
controls by age, gender and year of cohort entry from
2002 to 2008. The PUD patients were more likely to
suffer from the selected comorbidities and have a
higher rate of CCI scores greater than 1 based on
the dichotomized outcome of the Charlson scores.
Of the total 41 229 PUD patients, 2013 patients
experienced HZ during the follow-up period
(Table 1). The cumulative incidence of HZ in PUD
patients was significantly higher than that of the control cohorts (P < 0.001) over the follow-up period
(Figure 1). The incidence of HZ in PUD patients
(11.76 per 1000 patient-years) was significantly
greater than that in the control cohorts (6.56 per
1876
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Table 1 Demographic characteristics for adult patients with peptic ulcer disease and control cohorts in Taiwan
from 2002 to 2008
Patients with PUD
n ¼ 41 229
Number (%)
51.89 (16.28)
Variablea
Age (years), mean (SD)
Control cohorts
n ¼ 41 229
Number (%)
51.85 (16.24)
P*
0.74
0.99
Age group, years
18-24
1835 (4.45%)
1831 (4.44%)
25-34
4886 (11.85%)
4876 (11.83%)
35-44
8103 (19.65%)
8117 (19.69%)
45-54
9721 (23.58%)
9768 (23.69%)
55-65
6892 (16.72%)
6922 (16.79%)
^65
9792 (23.75%)
9715 (23.56%)
Female
19 353 (46.94%)
19 353 (46.94%)
Male
21 876 (53.06%)
21 876 (53.06%)
Gender
Follow-up (months), mean (SD)
50.53 (25.85)
52.88 (25.07)
1.00
<0.001
Comorbidities
Diabetes mellitus
5295 (12.84%)
Cancers
Hypertension
1991 (4.83%)
<0.001
1856 (4.50%)
546 (1.32%)
<0.001
10 641 (25.81%)
4371 (10.60%)
<0.001
32 864 (79.71%)
39 569 (95.97%)
<0.001
CCI score
0
^1
Herpes zoster, number (%)
8365 (20.29%)
1660 (4.03%)
2013 (4.88%)
1174 (2.85%)
<0.001
Controls
0.05
PUD
Log rank P-value<0.001
0.00
Proportion with herpes zoster
0.10
SD, standard deviation; NTD, New Taiwan Dollars; CCI, Charlson comorbidity index.
a
Chi-square tests for categorical variables; Student’s t test for continuous variables.
*P < 0.001, statistical significance.
0
Number at risk
PUD 41229
Controls 41229
2
4
Time (years)
6
8
33202
34506
20710
22135
9880
10728
0
0
Figure 1 Cumulative incidences of herpes zoster for adult
patients with peptic ulcer disease and the control cohorts
1000 patient-years). Cox regression revealed that PUD
was an independent predictor for HZ after adjusting
for potential confounders including age, gender, the
selected comorbidities and CCI scores [adjusted HR
1.77, 95% confidence interval (CI): 1.64-1.91;
P < 0.001) (Table 2).
Adjusted HRs for HZ in PUD patients not suffering
from any one of the selected comorbidities were significantly greater compared with those of the controls
(no individual comorbidity or PUD). In particular, the
adjusted HR of HZ in PUD patients with none of the
three selected comorbidities was 1.94 (95% CI: 1.772.12) compared with that of the controls after adjusting for potential confounders (P < 0.001) (Table 3).
These results suggest that PUD is associated with an
increased risk of developing HZ, independent of other
risk factors (selected comorbidities and CCI).
Based on the risk factors of PUD, stratified analyses
revealed that the adjusted HRs for HZ compared with
the control cohorts were ranked from high to low as:
the combination of nsNSAID usage and H. pylori infection (2.19, 95% CI: 1.96-2.42); nsNSAID usage only
(2.14, 95% CI: 1.95-2.37); H. pylori infection only
(1.40, 95% CI: 1.24-1.58); and Others11 (1.34, 95%
CI: 1.19-1.50) (Table 4). Our results indicated that
any risk factors in the four subgroups of PUD predisposed adult PUD patients to HZ; however, the
PEPTIC ULCER DISEASE INCREASES ZOSTER INCIDENCE
1877
Table 2 Incidence and hazard ratios for herpes zoster during the follow-up period for adult patients with peptic ulcer
disease versus control cohorts
All subjects
(n ¼ 90 846)
PUD
Control
HZ
2013
Persons
at risk
171 114
Incidence per
1000 person-years
11.76
1174
179 087
6.56
Adjusted HRa (95% CI)
1.77* (1.64–1.91)
1.00
HZ, herpes zoster;HR, hazard ratio.
a
Adjustments were made for age, gender, and selected comorbidities (diabetes mellitus, cancers, and hypertension) and the
Charlson comorbidity index.
*P < 0.001 for statistical significance.
Table 3 Incidence and hazard ratios for adult patients with peptic ulcer disease and control cohorts concomitant with
comorbidities
PUD patients
Comorbidity
Cases
Control cohorts (No PUD)
HZ
Person
at risk
Incidence
per 1000
person-years
Cases
HZ
Person
at risk
Incidence
per 1000
person-years
Adjusted HRa
(95% CI)
P
Diabetes mellitus
Yes
5295
316
19 555
16.16
1991
98
7963
12.31
1.39 (1.10–1.75)
0.006
No
35 934
1697
151 558
11.20
39 238
1076
171 125
6.29
1.80 (1.67–1.95)
<0.001
Cancers
Yes
1856
105
5360
19.59
546
25
2031
12.31
1.68 (1.07–2.62)
0.023
No
39 373
1908
165 754
11.51
40 683
1149
177 056
6.49
1.77 (1.64–1.92)
<0.001
<0.001
Hypertension
Yes
10 641
646
40 560
15.93
4371
209
17 777
11.76
1.42 (1.21–1.67)
No
30 588
1367
130 554
10.47
36 858
965
161 311
5.98
1.86 (1.70–2.02)
<0.001
27 576
1205
11 9588
10.08
35 541
888
156 022
5.69
1.94 (1.77–2.12)
<0.001
None of above
PUD, peptic ulcer disease; HR, hazard ratio; HZ, herpes zoster.
a
Adjusted by age, gender, selected comorbidities and the Charlson comorbidity index.
Table 4 Independent predictors of herpes zoster identified
by Cox hazard regressions in adult patients with peptic ulcer
disease
Risk factors of PUD
Control cohorts (No PUD)
Adjusted HRa
(95% CI)
1.00
þ H. pylori þ nsNSAID
2.19* (1.96–2.42)
nsNSAID only (- H. pylori; þ nsNSAID)
2.14* (1.95–2.37)
H. pylori only (þ H. pylori; - nsNSAID)
1.40* (1.24–1.58)
Others (- H. pylori; - nsNSAID)
1.34* (1.19–1.50)
PUD, peptic ulcer disease; nsNSAID, non-selective nonsteroidal
anti-inflammatory drugs; Others containing usage of cyclooxygenase-2 inhibitors, aspirin, non-aspirin anti-coagulants and
other ulcerogenic medications.
a
Adjusted hazard ratio: adjusted by age group, gender, and
three-selected comorbidities.
*P < 0.001 for statistical significance.
increased incidence of HZ varied with the different
risk factors of PUD. PUD associated with the combination of nsNSAID usage and H. pylori infection put
patients at the greatest risk of HZ.
Over time, a decreased number of patients with HZ
among all adult PUD patients annually was noted,
from 2002 to 2008. However, the annual incidence
of HZ in all PUD patients (ranging from 11.14 to
12.92) did not change during the study periods
(Table 5).
Discussion
To our knowledge, this is the first population-based
cohort study to assess the risk of HZ in adult populations with PUD. In this study, adults with PUD were
found to be at greater risk for developing HZ than the
control cohorts. Annual incidence of HZ in all PUD
patients did not change during the study periods.
Stratified analyses by risk factors of PUD revealed
that PUD including any risk factors of PUD subgroups
predisposed adult patients to HZ. PUD associated with
the combination of nsNSAID usage and H. pylori infection put patients at the greatest risk of HZ in stratified analyses.
1878
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Table 5 Numbers of herpes zoster patients and incidence for herpes zoster in all adult patients with
peptic ulcer disease (PUD) during the follow-up period from 2002 to 2008
PUD patient
numbers
6020
Herpes zoster
numbers
475
Persons
at risk
42 634
Incidence
per 1000
person-years
11.14
2003
5089
384
30 960
12.40
2004
5999
353
30 888
11.43
2005
5611
298
23 071
12.92
2006
5798
217
19 067
11.38
2007
6559
177
15 578
11.36
2008
6153
109
8917
12.22
Calendar year
2002
HZ can cause acute pain and long-term morbidities,
causing a negative impact on patients’ quality of life.
Medical costs associated with HZ and zoster-related
complications place a substantial burden on the
healthcare system.1,3 An HZ vaccine for adults has
been available since 2006. The present zoster vaccine
guidelines, based primarily on age, recommend administering HZ vaccine to immunocompetent individuals
aged 60 years or older. However, uptake rates of
zoster vaccination are low.6,7 Much research concerning comorbidities has surveyed possible predictors of
HZ in order to improve HZ vaccination strategies.8,25
PUD is an important health concern because of its
high prevalence (4–17%) in adult populations worldwide.31–33 In this study, adult PUD patients were 1.77
times more likely to develop HZ than the general population. Our results suggest that PUD should be added
to the list of risk factors for developing HZ.
Furthermore, adult PUD patients were found more
likely to suffer from the selected comorbidities and
CCI scores greater than 1 than the control cohorts.
Pre-specified subgroup analyses on exposure to the selected comorbidities showed that the adjusted HRs of
HZ in PUD patients with none of the selected comorbidities was 1.94 (95% CI: 1.77–2.12) as compared with
the controls (P < 0.001). The data indicate that PUD is
associated with an increased risk for developing HZ
after excluding the potential confounding effects of
comorbidities. Furthermore, our findings in subgroup
analyses concerning comorbidities are consistent with
previous studies that have identified diabetes mellitus,
cancers and hypertension25 as risk factors of HZ. Our
results suggest that the benefits for HZ vaccination
need to be evaluated in patients with PUD and other
aforementioned comorbidities in future studies.
Major risk factors of PUD are nsNSAID usage and/or
H. pylori infection.11 Stratified analyses by the risk factors of PUD in PUD patients showed that the adjusted
HRs for HZ in the two nsNSAID-related subgroups
were higher than those in the other two subgroups
including H. pylori infection only and Others.
Although adjusted HRs of the fourth subgroup
(Others) including usage of cyclooxygenase-2
inhibitors and other ulcerogenic medications increased
the risk of HZ significantly, adjusted HRs of the fourth
subgroup (Others) ranked the lowest among the four
subgroups of PUD. The findings are consistent with
previous studies showing that current exposures to
NSAIDs, including nsNSAID and cyclooxygenase-2 inhibitors, increase the risk of HZ and zoster-related
complications in human observational studies.10,34
However, our data revealed different influences on
the risk of HZ between nsNSAID usage and cyclooxygenase-2 inhibitors/other ulcerogenic medications
usage. The major morbidity caused by HZ is acute herpetic and chronic pain. NSAIDs, opioids, tricyclic antidepressants and gabapentin are capable of reducing
herpetic pain. Prescribing NSAIDs for treating herpetic
pain was common in a population-based study.4
Therefore, our data suggest that treating herpetic
pain with NSAIDs, especially nsNSAIDs, should be
avoided in PUD patients. First, immunity to varicella
zoster virus may be decreased by nsNSAIDs exposure.6,10 Second, symptoms and signs of PUD may be
worsened by exposure to nsNSAIDs because of their
ulcerogenic properties. Based on our study and previous studies, HZ vaccination could be emphasized in
PUD patients, especially those using nsNSAIDs.
Several potential explanations can be made for the
possible underlying mechanisms of the association between HZ and PUD. The risk of HZ is a comprehensive
result of cellular immunity that has been linked to the
ageing process2,21 and to nutritional status.21,22,35 PUD
has been linked to immunological dysfunctions14,15
and nutritional deficiencies13,16,19,20 including proteins,
vitamins A and C and zinc. Nutritional deficiencies13,20,36,37 may decrease specific immune responses
to varicella zoster virus.38–40 As a consequence, PUD
patients could be prone to developing HZ. In addition,
nutritional deficiencies of vitamin C and zinc have
been identified as predictors for HZ35,41–43 and postherpetic neuralgia (PHN).44,45 Thus, it is reasonable that
our results suggest that PUD patients are associated
with an increased risk of HZ through the possible
common pathways of impaired cellular immunity
and/or depressed nutritional status.
PEPTIC ULCER DISEASE INCREASES ZOSTER INCIDENCE
Strengths
The main strength of our study is that our data were
obtained from non-biased representative samples
from a nationwide population-based database, in
which up to 96% of Taiwan’s population was enrolled.9 An additional strength of our study is the
relative homogeneity of Taiwan’s population, which
reduces potential confounding by race, because
racial differences may be a risk factor for developing
HZ.46 Furthermore, patients with HZ received their
diagnosis either in inpatient or outpatient clinics,
which lessened the chance of selection bias arising
from including only severely affected patients. Data
on death were also obtained and calculated in this
analysis. Thus, our data including person-years at
risk and incidence per 1000 person-years were less
biased than the results in previous studies.9
Limitations
First, this study is based on a database. Cases of HZ
were defined by diagnostic codes without medical
record review.47 Second, H. pylori-associated PUD
was defined as the usage of H. pylori eradication therapy in our study. It is impossible to identify those
who did not receive H. pylori eradication therapy, in
PUD patients and the control cohorts. Third, it is not
possible for this study to clarify all of the causal relationships among PUD, HZ and other comorbidities.
Fourth, our database did not include nutritional and
immunological data of patients such as varicella
voster virus (VZV)-specific T cell-mediated immunity.
Nevertheless, the negative influence of PUD on the
incidence of HZ is validated in the present study.
Fifth, all patients in this study were Taiwanese and
lived in a tropical region. Racial groups and geographical regions have been found to be independent factors in HZ and postherpetic neuralgia.48 It is possible
that our results may not be applicable to the worldwide population. Thus, future large-scale investigations into other racial groups and geographical
regions are needed to confirm the findings.
In conclusion, to our knowledge, this populationbased cohort study is the first report to assess the
risk of HZ in an adult population with PUD. In this
1879
study, PUD is found to be an independent risk factor
for HZ. The PUD patients are more in need of vaccination than the general population, to lower incidence
of HZ. Further studies are required to clarify the
nature of the underlying associations between HZ
and PUD.49
Supplementary Data
Supplementary data are available at IJE online.
Funding
The present study was supported by grants
(CMNDMC9712 and CMFHR10080) provided by the
Chi Mei Medical Center, Tainan, Taiwan.
Acknowledgements
We appreciate the helpful remarks of Chung-Yi Li,
Professor and Chairman of the Department of Public
Health, College of Medicine, National Cheng Kung
University, Tainan, Taiwan. We appreciate the suggestion of the reviewer to increase the validity of the
positive findings in this observational study by using
falsification endpoints. These can be found in the supplemental file (available as Supplementary data at IJE
online).
Specific author contributions
Drafting the manuscript: Jen-Yin Chen; performing
statistical analysis: Shih-Feng Weng; all authors
have been involved in critical study concepts and
design, acquisition of data, revisions and editing of
the manuscript and have approved the final draft;
study supervision: Miao-Lin Hu. Jen-Yin Chen (the
first author) will act as guarantor for the paper.
Conflict of interest: None declared.
KEY MESSAGES
In this study, adult PUD patients were 1.77 times more likely to develop herpes zoster than the
general population. This suggests that adult PUD patients are more in need of herpes zoster vaccination than the general population.
Stratified analyses by the risk factors of PUD showed that any risk factors in the four subgroups of
PUD predispose adult PUD patients to herpes zoster. Interestingly, adjusted HRs for herpes zoster in
the two nsNSAID-related subgroups were higher than those in the other two subgroups including
H. pylori infection only and Others.
Our data indicate that the increased incidence of herpes zoster varied with the different risk factors of
PUD.
1880
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Yawn BP, Itzler RF, Wollan PC, Pellissier JM, Sy LS,
Saddier P. Health care utilization and cost burden of
herpes zoster in a community population. Mayo Clin Proc
2009;84:787–94.
Oxman MN, Levin MJ, Johnson GR et al. A vaccine to
prevent herpes zoster and postherpetic neuralgia in older
adults. N Engl J Med 2005;352:2271–84.
Yawn BP, Saddier P, Wollan PC, St Sauver JL,
Kurland MJ, Sy LS. A population-based study of the incidence and complication rates of herpes zoster before
zoster vaccine introduction. Mayo Clin Proc 2007;82:
1341–49.
Jih JS, Chen YJ, Lin MW et al. Epidemiological features
and costs of herpes zoster in Taiwan: a national study
2000 to 2006. Acta Derm Venereol 2009;89:612–16.
Lin YH, Huang LM, Chang IS et al. Disease burden and
epidemiology of herpes zoster in pre-vaccine Taiwan.
Vaccine 2010;28:1217–20.
Rimland D, Moanna A. Increasing incidence of herpes
zoster among Veterans. Clin Infect Dis 2010;50:1000–05.
Opstelten W, van Essen GA, Hak E. Determinants of noncompliance with herpes zoster vaccination in the community-dwelling elderly. Vaccine 2009;27:192–96.
Joesoef RM, Harpaz R, Leung J, Bialek SR. Chronic medical conditions as risk factors for herpes zoster. Mayo Clin
Proc 2012;87:961–67.
Yang YW, Chen YH, Wang KH, Wang CY, Lin HW. Risk of
herpes zoster among patients with chronic obstructive
pulmonary disease: a population-based study. Cmaj
2011;183:E275–80.
Mikaeloff Y, Kezouh A, Suissa S. Nonsteroidal antiinflammatory drug use and the risk of severe skin and
soft tissue complications in patients with varicella or
zoster disease. Br J Clin Pharmacol 2008;65:203–09.
Ramakrishnan K, Salinas RC. Peptic ulcer disease. Am
Fam Physician 2007;76:1005–12.
Wu CY, Wu MS, Kuo KN, Wang CB, Chen YJ, Lin JT.
Long-term peptic ulcer rebleeding risk estimation in patients undergoing haemodialysis: a 10-year nationwide
cohort study. Gut 2011;60:1038–42.
Salgueiro J, Zubillaga M, Goldman C et al. Review article:
is there a link between micronutrient malnutrition and
Helicobacter pylori infection? Aliment Pharmacol Ther
2004;20:1029–34.
Gerhard M, Schmees C, Voland P et al. A secreted lowmolecular-weight protein from Helicobacter pylori induces cell-cycle arrest of T cells. Gastroenterology 2005;
128:1327–39.
Satoh Y, Ogawara H, Kawamura O, Kusano M,
Murakami H. Clinical Significance of Peripheral Blood T
Lymphocyte Subsets in Helicobacter pylori-Infected
Patients. Gastroenterol Res Pract 2012;2012:819–42.
Malfertheiner P, Megraud F, O’Morain C et al. Current
concepts in the management of Helicobacter pylori infection: the Maastricht III Consensus Report. Gut 2007;56:
772–81.
Cho JY. Immunomodulatory effect of nonsteroidal antiinflammatory drugs (NSAIDs) at the clinically available
doses. Arch Pharm Res 2007;30:64–74.
Girndt M, Sester U, Kaul H, Hunger F, Kohler H.
Glucocorticoids inhibit activation-dependent expression
of costimulatory molecule B7-1 in human monocytes.
Transplantation 1998;66:370–75.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Bjarnason I, Hayllar J, MacPherson AJ, Russell AS. Side
effects of nonsteroidal anti-inflammatory drugs on the
small and large intestine in humans. Gastroenterology
1993;104:1832–47.
Nair S, Norkus EP, Hertan H, Pitchumoni CS.
Micronutrient antioxidants in gastric mucosa and serum
in patients with gastritis and gastric ulcer: does
Helicobacter pylori infection affect the mucosal levels?
J Clin Gastroenterol 2000;30:381–85.
Lesourd B. Nutritional factors and immunological ageing.
Proc Nutr Soc 2006;65:319–25.
Maggini S, Wintergerst ES, Beveridge S, Hornig DH.
Selected vitamins and trace elements support immune
function by strengthening epithelial barriers and cellular
and humoral immune responses. Br J Nutr 2007;
98(Suppl 1):S29–35.
Hsiao FY, Tsai YW, Wen YW, Kuo KN, Tsai CR,
Huang WF. Effect of Helicobacter pylori eradication therapy on risk of hospitalization for a major ulcer event.
Pharmacotherapy 2011;31:239–47.
Bureau of National Health Insurance in Taiwan. http://
www.nhi.gov.tw/english/index.aspx.
Hata A, Kuniyoshi M, Ohkusa Y. Risk of Herpes zoster in
patients with underlying diseases: a retrospective hospital-based cohort study. Infection 2011;39:537–44.
Wang FW, Tu MS, Mar GY et al. Prevalence and risk
factors of asymptomatic peptic ulcer disease in Taiwan.
World J Gastroenterol 2011;17:1199–203.
Wang YP, Liu CJ, Hu YW, Chen TJ, Lin YT, Fung CP. Risk
of cancer among patients with herpes zoster infection: a
population-based study. Cmaj 2012;184:E804–09.
Kuo CC, Lee CT, Lee IM, Ho SC, Yang CY. Risk of herpes
zoster in patients treated with long-term hemodialysis: a
matched cohort study. Am J Kidney Dis 2012;59:428–33.
Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical
comorbidity index for use with ICD-9-CM administrative
databases. J Clin Epidemiol 1992;45:613–19.
Greenland S, Pearl J, Robins JM. Causal diagrams for
epidemiologic research. Epidemiology 1999;10:37–48.
Bernersen B, Johnsen R, Straume B, Burhol PG,
Jenssen TG, Stakkevold PA. Towards a true prevalence
of peptic ulcer: the Sorreisa gastrointestinal disorder
study. Gut 1990;31:989–92.
Aro P, Storskrubb T, Ronkainen J et al. Peptic ulcer disease in a general adult population: the Kalixanda study: a
random population-based study. Am J Epidemiol 2006;
163:1025–34.
Li Z, Zou D, Ma X et al. Epidemiology of peptic ulcer
disease: endoscopic results of the systematic investigation
of gastrointestinal disease in China. Am J Gastroenterol
2010;105:2570–77.
Wolfe F, Michaud K, Chakravarty EF. Rates and predictors of herpes zoster in patients with rheumatoid arthritis and non-inflammatory musculoskeletal disorders.
Rheumatology (Oxford) 2006;45:1370–75.
Thomas SL, Wheeler JG, Hall AJ. Micronutrient intake
and the risk of herpes zoster: a case-control study. Int J
Epidemiol 2006;35:307–14.
Milanino R, Frigo A, Bambara LM et al. Copper and zinc
status in rheumatoid arthritis: studies of plasma, erythrocytes, and urine, and their relationship to disease activity
markers and pharmacological treatment. Clin Exp
Rheumatol 1993;11:271–81.
Biondi C, Pavan B, Dalpiaz A, Medici S, Lunghi L,
Vesce F. Expression and characterization of vitamin C
PEPTIC ULCER DISEASE INCREASES ZOSTER INCIDENCE
38
39
40
41
42
43
transporter in the human trophoblast cell line HTR-8/
SVneo: effect of steroids, flavonoids and NSAIDs. Mol
Hum Reprod 2007;13:77–83.
Smith JG, Levin M, Vessey R et al. Measurement of cellmediated immunity with a Varicella-Zoster Virus-specific
interferon-gamma ELISPOT assay: responses in an elderly
population receiving a booster immunization. J Med Virol
2003;70(Suppl 1):S38–41.
Kuvibidila SR, Gardner R, Velez M, Yu L. Iron deficiency,
but not underfeeding reduces the secretion of interferongamma by mitogen-activated murine spleen cells. Cytokine
2010;52:230–37.
Sandstead HH, Prasad AS, Penland JG et al. Zinc deficiency
in Mexican American children: influence of zinc and other
micronutrients on T cells, cytokines, and antiinflammatory
plasma proteins. Am J Clin Nutr 2008;88:1067–73.
High KP, Legault C, Sinclair JA, Cruz J, Hill K, Hurd DD.
Low plasma concentrations of retinol and alpha-tocopherol in hematopoietic stem cell transplant recipients: the
effect of mucositis and the risk of infection. Am J Clin
Nutr 2002;76:1358–66.
Schencking M, Vollbracht C, Weiss G et al. Intravenous
vitamin C in the treatment of shingles: results of a multicenter prospective cohort study. Med Sci Monit 2012;18:
CR215–24.
Chen JY, Chang CY, Lin YS, Hu ML. Nutritional
factors in herpes zoster, postherpetic neuralgia,
44
45
46
47
48
49
1881
and zoster vaccination. Popul Health Manag 2012;15:
391–97.
Chen JY, Chang CY, Feng PH, Chu CC, So EC, Hu ML.
Plasma vitamin C is lower in postherpetic neuralgia patients and administration of vitamin C reduces spontaneous pain but not brush-evoked pain. Clin J Pain 2009;25:
562–69.
Chen JY, Chu CC, Lin YS, So EC, Shieh JP, Hu ML.
Nutrient deficiencies as a risk factor in Taiwanese patients with postherpetic neuralgia. Br J Nutr 2011;106:
700–07.
Schmader K, George LK, Burchett BM, Pieper CF,
Hamilton JD. Racial differences in the occurrence of
herpes zoster. J Infect Dis 1995;171:701–04.
Yawn BP, Wollan P, St Sauver J. Comparing
shingles incidence and complication rates from
medical record review and administrative database
estimates: how close are they? Am J Epidemiol 2011;174:
1054–61.
Borkar DS, Tham VM, Esterberg E et al. Incidence of
herpes zoster ophthalmicus: results from the pacific
ocular inflammation study. Ophthalmology 2013;120:
451–56.
Chen JY, Chang CY, Lan KM, Sheu MJ, Lu CL, Hu ML. Is
peptic ulcer disease a risk factor of postherpetic neuralgia
in patients with herpes zoster? Med Hypotheses 2013; doi:
10.1016/j.mehy.2013.09.007.