International Journal of

Environmental Research
and Public Health

Article
Colorectal Cancer Risk in Patients with Hemorrhoids: A 10-Year
Population-Based Retrospective Cohort Study
En-Bo Wu 1 , Fung-Chang Sung 2,3,4 , Cheng-Li Lin 3 , Kuen-Lin Wu 5 and Kuen-Bao Chen 1,6, *

1 Department of Anesthesiology, China Medical University Hospital, China Medical University,

Taichung 404, Taiwan; [email protected]
2 Department of Health Services Administration, College of Public Health, China Medical University,
Taichung 404, Taiwan; [email protected]
3 Management Office for Health Data, China Medical University Hospital, Taichung 404, Taiwan;
[email protected]
4 Department of Food Nutrition and Health Biotechnology, Asia University, Taichung 413, Taiwan
5 Division of Colorectal Surgery, Department of Surgery, Kaohsiung Chang-Gung Memorial Hospital and
Chang-Gung University College of Medicine, Kaohsiung 833, Taiwan; [email protected]
6 Department of Anesthesiology, College of Medicine, China Medical University, Taichung 404, Taiwan
* Correspondence: [email protected]

Abstract: Colorectal cancer (CRC) is a common disease and one of the leading causes of cancer deaths
worldwide. This retrospective cohort study evaluated the risk of developing CRC in people with
hemorrhoids. Using Taiwan’s National Health Insurance Research Database, we established three sets





 of retrospective study cohorts with and without hemorrhoids. The first set of cohorts were matched
Citation: Wu, E.-B.; Sung, F.-C.; Lin,
by sex and age, the second set of cohorts were matched by propensity score without including
C.-L.; Wu, K.-L.; Chen, K.-B. colonoscopies, and the third set of cohorts were matched by propensity score with colonoscopies,
Colorectal Cancer Risk in Patients colorectal adenomas, and appendectomies included. In the second set of cohorts, 36,864 persons
with Hemorrhoids: A 10-Year with hemorrhoids that were diagnosed from 2000 to 2010 and a comparison cohort, with the same
Population-Based Retrospective size and matched by propensity score, were established and followed up to the end of 2011 to assess
Cohort Study. Int. J. Environ. Res. the incidence and Cox proportional regression-measured hazard ratio (HR) of CRC. The overall
Public Health 2021, 18, 8655. https:// incidence rate of CRC was 2.39 times greater in the hemorrhoid cohort than it was in the comparison
doi.org/10.3390/ijerph18168655
cohort (1.29 vs. 0.54 per 1000 person-years), with a multivariable model measured adjusted HR
of 2.18 (95% CI = 1.78–2.67) after controlling for sex, age, and comorbidity. Further analysis on
Academic Editors: Kelly A. Hirko,
the CRC incidence rates among colorectal sites revealed higher incidence rates at the rectum and
Dorothy R. Pathak and Paul
sigmoid than at other sites, with adjusted HRs 2.20 (95% CI = 1.48–3.28) and 1.79 (95% CI = 1.06–3.02),
B. Tchounwou
respectively. The overall incidence rates of both cohorts were similar in the first and second sets of
Received: 18 May 2021 cohorts, whereas the rate was lower in the third set of hemorrhoid cohorts than in the respective
Accepted: 13 August 2021 comparison cohorts, probably because of overmatching. Our findings suggest that patients with
Published: 16 August 2021 hemorrhoids were at an elevated risk of developing CRC. Colonoscopy may be strongly suggested
for identifying CRC among those with hemorrhoids, especially if they have received a positive fecal
Publisher’s Note: MDPI stays neutral occult blood test result.
with regard to jurisdictional claims in
published maps and institutional affil- Keywords: hemorrhoid; colorectal cancer; risk factor; propensity score matching; retrospective
iations. cohort study

Copyright: © 2021 by the authors. 1. Introduction

Licensee MDPI, Basel, Switzerland. Colorectal cancer (CRC) is a common disease and one of the leading causes of cancer
This article is an open access article deaths worldwide. Both genetic and environmental factors are associated with the risk of
distributed under the terms and
developing CRC. The consumption of red meat, lack of physical activity, obesity, cigarette
conditions of the Creative Commons
smoking, and alcohol use are known risk factors of CRC [1]. Most CRCs arise from adeno-
Attribution (CC BY) license (https://
mas that progress to dysplasia and carcinoma. On average, the progression from adenoma
creativecommons.org/licenses/by/
to adenocarcinoma takes 10 years [2]. Neoplastic changes result from both inherited and
4.0/).

Int. J. Environ. Res. Public Health 2021, 18, 8655. https://doi.org/10.3390/ijerph18168655 https://www.mdpi.com/journal/ijerph
Int. J. Environ. Res. Public Health 2021, 18, 8655 2 of 10

acquired genetic defects. Globally, CRC is the third most commonly diagnosed malignancy
in males and second in females, and the incidences vary markedly in the World Health
Organization GLOBOCAN database [3].
Hemorrhoids are common gastrointestinal disorders that are diagnosed by general
practitioners. The disease is characterized by inflammation of the anorectum, including
the submucosal, fibrovascular, and arteriovenous sinusoids [4]. Painless rectal bleeding
is a common complaint during defecation, particularly when tissue prolapse appears.
Other typical symptoms include anal pruritus, pain, and a lump at the anal verge due
to thrombosis or strangulation [5]. A low-fiber diet and low water intake; increased
intra-abdominal pressure caused by pregnancy, constipation, or prolonged straining; and
weakened muscular support are risks factors for developing hemorrhoids [6,7]. The clinical
manifestations of hemorrhoids are diverse, ranging from asymptomatic to rectal bleeding.
Most people who have hemorrhoids are free of symptoms and usually do not require
treatment [8].
CRC and hemorrhoids share several risk factors, such as low fiber intake, obesity, and
lack of adequate exercise [9]. Hemorrhoids also present symptoms that are comparable
to those of CRC, in particular, the presence of blood in the patient’s stool. Whether
hemorrhoids are associated with the development of CRC is a known concern among the
public, and consequently, screening for CRC among patients with hemorrhoids is often
advisable as a precautionary measure. However, few studies have actually examined
the relationship between CRC and hemorrhoids. Therefore, we conducted a population-
based retrospective cohort study to explore the link between hemorrhoids and CRC using
longitudinal insurance claims data from Taiwan.

2. Materials and Methods

2.1. Database
We obtained data from the National Health Research Institutes of Taiwan’s Longitudi-
nal Health Insurance Database (LHID), which contains claims data of 1 million insured
people that were randomly selected from 23 million registered users of the National Health
Insurance (NHI) system for the period from 1996 to 2011. Information on patients’ demo-
graphics, health care received, and health care cost is available in the claims data. Diseases
that were registered in the claims data can be identified using the International Classifi-
cation of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). The identification
numbers of those included in the data were scrambled with surrogate numbers to protect
privacy. This study was approved by the Ethics Review Committee of China Medical
University and Hospital (CMUH104-REC2-115-CR5).

2.2. Study Population and Comorbidity

From the LHID, we identified adult patients with hemorrhoids (ICD-9-CM 455) that
were newly diagnosed from 2000 to 2010 as the hemorrhoid cohort. The first diagnosis date
was considered as the index date. In order to delineate the role of a colonoscopy in CRC
detection, we established three sets of the comparison cohorts without hemorrhoid histories
from the LHID. In the first set, we randomly selected a comparison cohort with a sample
size that was fourfold the size of the hemorrhoid cohort and was frequency matched by sex
and age. In the second and third sets of cohorts, we established a comparison cohort with a
sample size similar to the size of the hemorrhoid cohort, which was frequency matched
by propensity score. The propensity score was estimated for each insured resident from a
logistic regression with sociodemographic and potential comorbidities that are associated
with cancer risk as covariates, including inflammatory bowel disease (IBD) (ICD-9-CM
555 and 556), hypertension (ICD-9-CM 401-405), diabetes (ICD-9-CM 250), hyperlipidemia
(ICD-9-CM 272), stroke (ICD-9-CM 430-438), congestive heart failure (ICD-9-CM 428),
obesity (ICD-9-CM 278), pyogenic liver abscess (PLA) (ICD-9-CM 572.0), hepatitis B virus
(HBV) (ICD-9-CM V02.61, 070.20, 070.22, 070.30, and 070.32), hepatitis C virus (HCV) (ICD-
9-CM V02.62, 070.41, 070.44, 070.51, 070.54, and 070.70-71), chronic obstructive pulmonary
Int. J. Environ. Res. Public Health 2021, 18, x 3 of 10

Int. J. Environ. Res. Public Health 2021, 18, 8655 3 of 10

hepatitis C virus (HCV) (ICD-9-CM V02.62, 070.41, 070.44, 070.51, 070.54, and 070.70-71),
chronic obstructive pulmonary disease (COPD) (ICD-9-CM 491-492, and 496), alcohol-re-
disease
lated (COPD)
illness (ICD-9-CM
(ICD-9-CM 491-492,
V11.3, A215, 291,and
303,496),
305, alcohol-related illness
571.0-3, and 790.3) and(ICD-9-CM V11.3,
chronic pancre-
A215, 291, 303, 305, 571.0-3, and 790.3) and chronic pancreatitis (ICD-9-CM 577.1). The
atitis (ICD-9-CM 577.1). The histories of coloscopy use, colorectal adenomas, and appen-
histories of coloscopy use, colorectal adenomas, and appendectomy were also included in
dectomy were also included in the propensity score estimation for the third set of study
the propensity score estimation for the third set of study cohorts, but not for the second set
cohorts, but not for the second set of cohorts.
of cohorts.
2.3. Statistical
2.3. Statistical Analysis
Analysis
WeWe consideredthe
considered thesecond
secondset setof of study
study cohorts
cohorts to to have
have had hadan anappropriate
appropriatedesign designbe-
because the third set of study cohorts may have been overmatched
cause the third set of study cohorts may have been overmatched by including colonoscopy, by including colonos-
copy, colorectal
colorectal adenomas,
adenomas, and and appendectomy
appendectomy as covariates
as covariates in the inestimation
the estimationof the ofpropensity
the pro-
pensity
score. score.
Data
Data analyses
analyses first
firstcompared
comparedthe thecharacteristics
characteristicsofofstudy
studysubjects
subjectsbetween
between the the hem-
hemor-
orrhoid and comparison cohorts. The distributions of categorical
rhoid and comparison cohorts. The distributions of categorical variables were examined variables were examined
using
usingchi-squared
chi-squared testing
testing andandthethe
means
means were
wereexamined
examined using
usingStudent’s
Student’s t-test. WeWe
t-test. esti-
esti-
mated and plotted the cumulative incidence of CRC 1 year after
mated and plotted the cumulative incidence of CRC 1 year after the index date using the index date using thethe
Kaplan–Meier
Kaplan–Meier method
method andandtested
tested thethe
difference
difference between
between curves
curves using
using the
the log-rank
log-rank test
test
(Figure
(Figure1).1).TheTheincidence
incidence density
density rates
ratesof of
CRCCRC were
were also calculated
also calculated 1 year
1 year after
afterthe index
the index
date.
date.WeWe used
used Cox
Cox proportional
proportional hazards
hazards regression
regression analysis
analysis toto
calculate
calculate thethehazard
hazard ratio
ratio
(HR)
(HR)andandthe the corresponding 95% 95%confidence
confidenceinterval
interval (CI)(CI) of hemorrhoid
of the the hemorrhoid cohortcohort
against
against the comparison
the comparison cohort. cohort. Multivariable
Multivariable analysisanalysis
was used was to used to estimate
estimate the adjusted
the adjusted hazard
hazard ratio (aHR). Overall incidence and incidence by colorectum
ratio (aHR). Overall incidence and incidence by colorectum site were also estimated, site were also esti-in-
mated,
cludingincluding in the ascending
in the ascending colon (ICD-9-CM
colon (ICD-9-CM 153.0 and153.0 and transverse
153.4-6), 153.4-6), transverse colon
colon (ICD-9-CM
(ICD-9-CM 153.1), descending
153.1), descending colon (ICD-9-CM
colon (ICD-9-CM 153.2 and153.2153.7),and 153.7), colon
sigmoid sigmoid colon (ICD-9-
(ICD-9-CM 153.3),
CM 153.3),
rectum rectum (ICD-9-CM
(ICD-9-CM 154.0-1), and 154.0-1), and other
other locations locations 153.8-9,
(ICD-9-CM (ICD-9-CM 153.8-9,
154.2-3, 154.2-3,All
and 154.8).
and 154.8). All
statistical statistical
analysis analysis was
was performed using performed
SAS version using9.4 SAS version
software (SAS9.4 software
Institute, (SAS
Cary, NC,
USA), and
Institute, Cary,theNC,
figure of the
USA), andcumulative
the figureincidence curve wasincidence
of the cumulative plotted using
curveR wassoftware.
plotted The
significance
using R software. levelThewas set at p < 0.05
significance levelforwas
two-sided
set at p <testing.
0.05 for two-sided testing.

Figure
Figure 1. 1. Cumulative
Cumulative incidence
incidence of of colorectal
colorectal cancer
cancer of of cohorts
cohorts with
with and
and without
without hemorrhoids,
hemorrhoids, asas
obtained
obtained using
using thethe Kaplan–Meier
Kaplan–Meier method.
method.
Int. J. Environ. Res. Public Health 2021, 18, 8655 4 of 10

Characteristics of the study population of the first and third sets of cohorts are pre-
sented in Supplementary Tables S1 and S2. The overall incidence rates and HRs of CRC of
the three study sets are presented in Supplementary Table S3.

3. Results
In this section, we present the findings that were based on the data for the second set
of study cohorts, which had equal sample sizes of 36,864 persons.

3.1. Demography
These propensity-score-matched study cohorts were similar in prevalence rates of
most comorbidities, except that the rates of hypertension, diabetes, and congestive heart
failure were higher in the hemorrhoid cohort than in the comparison cohort (Table 1). More
than half of the study population were men and there were more in the hemorrhoid cohort
than there were in the comparison cohort. The hemorrhoid cohort was older than the
comparison cohort was.

Table 1. Demographic characteristics and comorbidities that were compared between cohorts with
and without hemorrhoids who were matched by propensity score without colonoscopies.

Hemorrhoids
Standardized
Variable No Yes
n = 36,864 n = 36,864 Difference §

Sex n (%) n (%)

Female 17,582 (47.7) 16,809 (45.6) 0.04
Male 19,282 (52.3) 20,055 (54.4) 0.04
Age, mean (SD) 45.8 (15.9) 46.9 (15.8) 0.07
Stratified
≤49 years 23,569 (63.9) 22,622 (61.4) 0.05
50–64 7836 (21.3) 8443 (22.9) 0.04
65+ 5459 (14.8) 5799 (15.7) 0.03
Comorbidity
IBD 566 (1.54) 585 (1.59) 0.004
Hypertension 8580 (23.3) 9694 (26.3) 0.07
Diabetes 2120 (5.75) 2252 (6.11) 0.02
Hyperlipidemia 7206 (19.6) 7263 (19.7) 0.004
Stroke 830 (2.25) 894 (2.43) 0.01
Congestive heart failure 984 (2.67) 1116 (3.03) 0.02
Obesity 173 (0.47) 193 (0.52) 0.008
PLA 22 (0.06) 23 (0.06) 0.001
HBV 1512 (4.10) 1512 (4.10) 0.000
HCV 452 (1.23) 505 (1.37) 0.01
COPD 3879 (10.5) 3860 (10.5) 0.002
Alcohol-related illness 1395 (3.78) 1397 (3.79) 0.000
Chronic pancreatitis 46 (0.12) 62 (0.17) 0.01
§A standardized mean difference of ≤0.10 indicates a negligible difference between the two cohorts. SD, standard
deviation; IBD, inflammatory bowel disease; PLA, pyogenic liver abscess; HBV, hepatitis B virus; HCV, hepatitis
C virus; COPD, chronic obstructive pulmonary disease.

3.2. Higher Incidence Rates of CRC in Hemorrhoids Patients

Figure 1 shows that the cumulative incidence of colon cancer was 0.94% greater in the
hemorrhoid cohort than it was in the comparison cohort (log-rank test p < 0.001).

3.3. Subgroup Analysis of Demographic Aspects and Comorbidities

After mean follow-up durations of 6.99 (SD = 3.12) and 6.85 years (SD = 3.13), the
incidence rate of CRC was 2.39-fold greater in patients with hemorrhoids than it was in
the comparison cohort, respectively (Table 2). After adjusting for age and gender, the
hemorrhoids cohort had an adjusted HR of 2.18 (95% CI = 1.78–2.67) for CRC. The CRC
Int. J. Environ. Res. Public Health 2021, 18, 8655 5 of 10

incidence increased with age and was higher in men than it was in women in both cohorts,
whereas the adjusted HR of CRC was slightly higher for women than it was for men.
Relative to the comparison group, the HRs of colon cancer were 3.53 (95% CI = 2.15–5.79),
1.91 (95% CI = 1.36–2.68), and 2.06 (95% CI = 1.53–2.77) for patients with hemorrhoids
aged ≤49, 50–64, and ≥65 years, respectively. Comorbidity was also associated with an
excess increase in CRC incidence, which was greater in the hemorrhoid cohort than in the
comparison cohort (1.49 versus 0.56 per 1000 person-years) compared to those without
comorbidity.

Table 2. Incidence and hazard ratio of colorectal cancer compared between the cohorts with and without hemorrhoids by
gender, age, and comorbidity.

Hemorrhoids
No Yes Crude HR Adjusted HR
Variable
(95% CI) (95% CI)
Event PY Rate # Event PY Rate #
All 138 255,722 0.54 337 261,466 1.29 2.39 (1.96, 2.93) * 2.18 (1.78, 2.67) *
Gender
Female 52 113,232 0.46 128 121,173 1.06 2.35 (1.71, 3.25) * 2.19 (1.59, 3.03) *
Male 86 142,490 0.60 209 140,293 1.49 2.38 (1.84, 3.09) * 2.16 (1.66, 2.80) *
Age-
stratified
≤49 23 162,628 0.14 71 165,694 0.43 3.53 (2.15, 5.79) * 3.53 (2.15, 5.79) *
50–64 45 55,159 0.82 113 57,499 1.97 1.95 (1.39, 2.73) * 1.91 (1.36, 2.68) *
65+ 70 37,935 1.85 153 38,272 4.00 2.12 (1.58, 2.85) * 2.06 (1.53, 2.77) *
Comorbidity
No 54 161,577 0.33 98 149,717 0.65 2.58 (1.79, 3.72) * 2.53 (1.76, 3.66) *
Yes 84 94,145 0.89 239 111,749 2.14 2.13 (1.67, 2.72) * 2.09 (1.64, 2.67) *
Rate # : incidence rate per 1000 person-years; PY, person-years; HR, hazard ratio. Adjusted HR: estimated after controlling for sex, age, and
comorbidity. Comorbidity: patients with any one of the comorbidities (IBD, hypertension, diabetes, hyperlipidemia, stroke, congestive
heart failure, obesity, PLA, HBV, HCV, COPD, alcohol-related illness, and chronic pancreatitis) were assigned to the comorbidity group.
* p < 0.001.

3.4. Relevance of Tumor Site and Hemorrhoids

Table 3 presents the occurrence of CRC by site for both cohorts. The cancer incidence
was greater in the hemorrhoid cohort than it was in the comparison cohort for most sites,
particularly for cancer of the rectum and sigmoid colon, with adjusted HRs of 2.20 (95% CI
= 1.48–3.28) and 1.79 (95% CI = 1.06–3.02), respectively.

Table 3. Incidences and hazard ratio of colorectal cancer estimated by colorectum site compared between cohorts with and
without hemorrhoids.

Hemorrhoids
Crude HR Adjusted HR
Site of CRC No Yes (95% CI) (95% CI)
Event Rate # Event Rate #
Ascending colon (ICD-9-CM 153.0, 153.4-6) 13 0.05 21 0.08 1.49 (0.75, 2.97) 1.34 (0.67, 2.69)
Transverse colon (ICD-9-CM 153.1) 4 0.02 5 0.02 1.18 (0.32, 4.41) 1.08 (0.29, 4.07)
Descending colon (ICD-9-CM 153.2, 153.7) 3 0.01 10 0.04 3.13 (0.86, 11.4) 2.83 (0.78, 10.3)
Sigmoid colon (ICD-9-CM 153.3) 21 0.08 43 0.17 1.92 (1.14, 3.24) * 1.79 (1.06, 3.02) *
Rectum (ICD-9-CM 154.0-1) 34 0.14 86 0.33 2.40 (1.61, 3.57) ** 2.20 (1.48, 3.28) **
Other (ICD-9-CM 153.8-9, 154.2-3, 154.8) 18 0.07 39 0.15 2.08 (1.19, 3.64) 1.86 (1.06, 3.26)
Unknown 45 0.17 133 0.51 2.99 (1.97, 3.66) ** 2.76 (1.85, 3.67) **
Rate # : incidence rate per 1000 person-years; HR, hazard ratio. Adjusted HR: estimated after controlling for sex, age, and comorbidity.
* p < 0.05, ** p < 0.001.
Int. J. Environ. Res. Public Health 2021, 18, 8655 6 of 10

4. Discussion
This study established three sets of study cohorts. The sex- and age-matched hem-
orrhoid cohort had more prevalent baseline comorbidities than the comparison cohort
(Supplementary Table S1). The second and third sets of cohorts were propensity score
matched, which had similar distributions between hemorrhoid and comparison cohorts for
most comorbidities. We found that the overall CRC incidence rates were similar in the first
and second sets of cohorts, with 1.29 per 1000 person-years in the hemorrhoids cohorts
and 0.54 per 1000 person-years in the comparison cohorts (Table 2 and Supplementary
Table S3). In the third set of study cohorts, the CRC incidence was lower in the hemorrhoids
cohort than it was in the comparison cohort (1.22 versus 1.62 per 1000 person-years), with
an aHR of 0.80 (95% CI = 0.69, 0.94). We suspected that the third set of study cohorts
were overmatched by including colonoscopies, colorectal adenomas, and appendectomies
in the propensity score matching, in which both cohorts had similar prevalence rates of
colonoscopy uses at baseline. The prevalence of colorectal adenomas was even higher in
the hemorrhoid cohort than in the comparison cohort. Therefore, we considered the second
set of study cohorts as being adequate for studying the relationship between hemorrhoids
and CRC risk.
The advantage of this study was using a large data set to conduct a nationwide popula-
tion study with a long follow-up period [10]. The follow-up was started 12 months after the
index date to avoid the immortal effect. Therefore, the association between hemorrhoids
and the risk of CRC demonstrated in the present study was highly convincing [11].
CRC is a common malignancy that results in more than 600,000 deaths globally every
year [12]. Among the risk factors that are associated with this cancer, genetic factors account
for approximately 20% of all cases of the disease [13,14]. Other instances can be associated
with environmental causes rather than genetic factors. The typical pathophysiology of
CRC development has a time frame of a decade, with the polyp-to-adenocarcinoma stage
averaging 7–10 years [15]. Dysplastic adenomas are the most common form of premalignant
lesions [16]. Genetic features show that adenomatous polyposis coli gene mutations are
present in the early phase of CRC formation in approximately 70% of adenomas, with 49–
50% caused by KRAS mutations [17]. Subsequently, activated KRAS oncogene mutations
and inactivated mutations of the TP53 tumor suppressor gene can result in an adenoma–
carcinoma sequence. Comprehensive genome analyses have found APC, KRAS, TGFBR2,
SMAD4, and TP53 as driver genes with mutations frequently found in human colorectal
cancers [18,19].
To our knowledge, environmental factors and genetic factors can promote the for-
mation of CRC. Among various environmental factors, inflammation was implicated in
elevated cancer risk [20–22]. Studies have indicated the vital roles of IBD, Crohn’s dis-
ease, and ulcerative colitis in the development of CRC [23–25]. Studies also discovered
that upregulated interleukin-17, interleukin-23, and signal transducer and activator of
transcription results in tumor development in patients with intestinal inflammation [26,27].
However, a recent study has recommended low-dose aspirin for the chemoprevention
of CRC [28]. Instances in which the benefits outweigh the risks have not yet been clearly
defined for specific individuals. A prospective, double-blind, multidose, and placebo-
controlled clinical trial, namely, ASPIRED, proposed that acetylsalicylic acid may be related
to the inhibition of cyclooxygenase’s inflammatory mechanism and other inflammatory
biomarkers in the formation of CRC [29].
In an earlier case–control study in the United States, Tseng et al. reported that hemor-
rhoids, sexually transmitted diseases, physical inactivity, and multiple sexual partners are
factors that are associated with an increased risk of anal cancer [30]. Frequent receptive
anal intercourse is likely the physio-pathological cause that leads to inflammation and
cancer development. Information on sexual activity and physical inactivity is not available
in our data to evaluate these relationships. Another case–control study in San Francisco
also found the homosexual activity and hemorrhoids are associated with the development
of anal and rectal squamous cell carcinoma [31]. Long-lasting hemorrhoids may act sim-
Int. J. Environ. Res. Public Health 2021, 18, 8655 7 of 10

ilarly to chronic inflammatory dermatoses, which may result in regional inflammatory

reactions and evolve in the localized squamous-cell carcinomas. In our study, we also
found that hemorrhoids significantly increase the risk of distal colon cancer, including in
the sigmoid colon, rectum, anus, and other sites (Table 3). This finding implicates that
localized inflammation may lead to tumor formation. In addition, epigenetic factors may
also be associated with CRC risk. In a Swedish case–control study, Ungerbäck et al. exam-
ined polymorphisms in nuclear factor kappa B (NF-κB)-pathway-associated genes, which
are the major regulators of inflammation [32]. The results showed that advanced CRC is
associated with heterozygous and polymorphic TNFAIP3 (rs6920220), heterozygous NLRP3
(Q705K), and polymorphic NF-κB-94 ATTG ins/del genotypes. The proinflammatory NF-κB
pathway degenerates the tissue, increases the risk of sustained cell and DNA damage, and
promotes other tumor biomarkers that are activated in the primary and metastatic CRC.
These studies demonstrated that chronic inflammation is associated with the formation
of CRC.
To date, few large-scale studies have investigated the relationship between hemor-
rhoids and CRC. In our retrospective cohort study with a 12-year follow-up period, we
found that patients with hemorrhoids had an aHR of 2.18 for developing CRC after match-
ing comorbidities, some of which are proven risk factors for CRC [33], and adjusting
for age and gender. In a population-based retrospective cohort study in 2013, Lee et al.
observed that the presence of hemorrhoids was significantly associated with a vital risk
of developing CRC [34], as shown by an increased standardized incidence ratio (SIR) of
1.50. The SIR method has a limitation, namely, that the effects of comorbidities cannot
be analyzed. We contend that adjustment for confounding factors is crucial in big data
studies. The propensity score matching method creates a balanced dataset, allowing for
a simple and direct comparison of baseline covariates between experimental and control
groups. After matching the propensity score, it is used to derive a pseudo-randomized
dataset, which allows for unbiased estimation of the exposure effect. Therefore, matching
based on the propensity score is frequently used in the analysis of medical statistics. Our
study provides stronger evidence of a link between hemorrhoids and CRC after propensity
score matching and adjusting for associated risks. Our study also has the advantage of
estimating the risk variation between different colorectal sites. These results imply that
the location of colorectal cancer is consistent with our conclusions. However, whether the
shorter latent period for CRC with hemorrhoids is attributable to molecular changes is
unclear. Genetic or environmental dispositions may exist in Taiwanese patients and this
deserves further investigation.

5. Advantages and Limitations

This study was strengthened by using a longitudinal design with a high follow-up
rate and was based on population-based data with a reliable diagnosis. Taiwan’s LHID was
created to advance the health of all citizens. Health representatives from other countries
were interested in visiting with the Taiwan National Health Insurance Administration
(NHIA) to learn about the insurance program. It has been reported that 981 dignitaries
from 71 countries had visited the agency by the end of 2016 [10].
However, there are several limitations to our study. First, lifestyle information regard-
ing drinking, smoking, and diet, as well as genetic factors, were unavailable to observe the
related impacts. Second, the propensity score matching method was used to establish study
cohorts; the bias resulting from the retrospective nature of the study may have influenced
our findings. Third, all data involved were anonymous and patients were selected through
ICD codes; therefore, the accuracy of the diagnoses can be disputed. Fourth, patients with
hemorrhoids might be more likely to be screened for cancer and were more likely to be
diagnosed with the disease. We conducted further data analysis to compare the history
of colonoscopy examinations between cohorts with and without hemorrhoids. Results
showed that the hemorrhoid group had colonoscopies more frequently than the comparison
cohort had, particularly during the follow-up period (Supplementary Table S4). However,
Int. J. Environ. Res. Public Health 2021, 18, 8655 8 of 10

we found that hemorrhoid surgery could reduce the colorectal cancer incidence by 49.3%,
from 1.46 to 0.74 per 1000 person-years (Supplementary Table S5). We also found that the
hemorrhoid cohort without a hemorrhage of the rectum and/or anus (ICD-9-CM 569.3)
were at higher risk for colorectal cancer with an aHR of 2.09 (Supplementary Table S6).
These data may imply that hemorrhoids could be associated with a twofold increased risk
without the surgery. Therefore, the claim that hemorrhoids are associated with an increased
risk of CRC remains reliable.

6. Conclusions
Our propensity-score-matched retrospective study demonstrated that patients with
hemorrhoids were at a nearly twofold increased risk of developing CRC. Hemorrhoidec-
tomy is seen as a benefit to hemorrhoid patients since it provides a near 50% risk reduction
of subsequent CRC. CRC is one of the most critical findings of colonoscopies, which is
used for polypectomy and biopsy if any suspicious lesions are found. On the basis of these
findings, we propose that hemorrhoids are a specific risk factor for CRC. Consequently,
people with hemorrhoids should be encouraged to undergo a colonoscopy for the early
detection of CRC.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/10

.3390/ijerph18168655/s1, Table S1: Demographic characteristics and comorbidities in patients with
and without hemorrhoids who were matched by sex and age, Table S2: Demographic characteristics
and comorbidities in patients with and without hemorrhoids who were matched by propensity
score with colonoscopy included, Table S3: Overall incidence rates and hazard ratios of colorectal
cancer of the hemorrhoids and comparison cohorts in first, second, and third sets of cohorts, Table S4:
Number of colonoscopies received before and after baseline compared between cohorts with and
without hemorrhoids, Table S5: Incidences and hazard ratios of colorectal cancer associated with
hemorrhoid surgery, Table S6: Incidence and hazard ratios of colorectal cancer associated with polyp
or hemorrhage compared between the cohorts with and without hemorrhoids.
Author Contributions: E.-B.W. and F.-C.S. designed the study; C.-L.L. and F.-C.S. performed the
statistical analyses and generated figures and tables; E.-B.W. wrote the original draft of the manuscript;
K.-B.C., F.-C.S. and K.-L.W. provided intellectual input and corrected the manuscript; E.-B.W., K.-B.C.
and F.-C.S. were responsible for the final editing and review of the manuscript; K.-B.C. supervised
this study. All authors have read and agreed to the published version of the manuscript.
Funding: This study is supported in part by Taiwan Ministry of Health and Welfare Clinical Trial
Center (MOHW109-TDU-B-212-114004), MOST Clinical Trial Consortium for Stroke (MOST Clinical
Trial Consortium for Stroke (MOST 109-2321-B-039-002), Tseng-Lien Lin Foundation, Taichung,
Taiwan.
Institutional Review Board Statement: This study used insurance claims data with the approval of
the Research Ethics Committee (CMUH104-REC2-115-CR5). All the NHIRD data related to personal
identification were encrypted by the NHIA before being published. The confidentiality of subjects in
the dataset was protected by the NHIA regulations.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data are available from the NHIRD published by Taiwan NHIA. Due
to legal restrictions imposed by the government of Taiwan in relation to the “Personal Information
Protection Act,” the data cannot be made publicly available. Requests for data can be sent as a formal
proposal to the NHIRD (http://nhird.nhri.org.tw, accessed on 30 June 2021).
Acknowledgments: All authors would like to thank four anonymous reviewers and the editor for
their comments. This study was based in part on data from Taiwan’s NHIRD, as provided by the
Bureau of NHI, Department of Health, and managed by National Health Research Institutes. The
interpretation and conclusions contained herein do not represent those of NHIA, Department of
Health, or National Health Research Institutes.
Conflicts of Interest: The authors declare that there is no conflict of interest.
Int. J. Environ. Res. Public Health 2021, 18, 8655 9 of 10

References
1. Turner, N.D.; Lloyd, S.K. Association between red meat consumption and colon cancer: A systematic review of experimental
results. Exp. Biol. Med. 2017, 242, 813–839. [CrossRef] [PubMed]
2. Bailie, L.; Loughrey, M.B.; Coleman, H.G. Lifestyle risk factors for serrated colorectal polyps: A systematic review and meta-
analysis. Gastroenterology 2017, 152, 92–104. [CrossRef]
3. Ferlay, J.; Colombet, M.; Soerjomataram, I.; Mathers, C.; Parkin, D.M.; Piñeros, M.; Znaor, A.; Bray, F. Estimating the global cancer
incidence and mortality in 2018: GLOBOCAN sources and methods. Int. J. Cancer 2019, 144, 1941–1953. [CrossRef] [PubMed]
4. Jacobs, D. Hemorrhoids. N. Engl. J. Med. 2014, 371, 944–951. [CrossRef] [PubMed]
5. Lohsiriwat, V. Treatment of hemorrhoids: A coloproctologist’s view. World J. Gastroenterol. 2015, 21, 9245–9252. [CrossRef]
6. Sanchez, C.; Chinn, B.T. Hemorrhoids. Clin. Colon Rectal Surg. 2011, 24, 005–013. [CrossRef]
7. Sandler, R.S.; Peery, A.F. Rethinking what we know about hemorrhoids. Clin. Gastroenterol. Hepatol. 2019, 17, 8–15. [CrossRef]
[PubMed]
8. Migaly, J.; Sun, Z. Review of hemorrhoid disease: Presentation and management. Clin. Colon Rectal Surg. 2016, 29, 022–029.
[CrossRef]
9. Boyle, T.; Keegel, T.; Bull, F.; Heyworth, J.; Fritschi, L. Physical activity and risks of proximal and distal colon cancers: A systematic
review and meta-analysis. J. Natl. Cancer Inst. 2012, 104, 1548–1561. [CrossRef] [PubMed]
10. National Health Insurance Administration; Ministry of Health and Welfare, Taiwan. National Health Insurance Research
Database, Taiwan. Available online: https://nhird.nhri.org.tw/en/index.html (accessed on 30 May 2014).
11. Sherman, R.E.; Anderson, S.A.; Dal Pan, G.J.; Gray, G.W.; Gross, T.; Hunter, N.L.; LaVange, L.; Marinac-Dabic, D.; Marks, P.W.;
Robb, M.A.; et al. Real-world evidence—What is it and what can it tell us? N. Engl. J. Med. 2016, 375, 2293–2297. [CrossRef]
12. Gallimore, A.M.; Godkin, A. Epithelial barriers, microbiota, and colorectal cancer. N. Engl. J. Med. 2013, 368, 282–284. [CrossRef]
[PubMed]
13. Rustgi, A.K. The genetics of hereditary colon cancer. Genes Dev. 2007, 21, 2525–2538. [CrossRef]
14. Tenesa, A.; Dunlop, M. New insights into the aetiology of colorectal cancer from genome-wide association studies. Nat. Rev.
Genet. 2009, 10, 353–358. [CrossRef]
15. Brenner, H.; Kloor, M.; Pox, C.P. Colorectal cancer. Lancet 2014, 383, 1490–1502. [CrossRef]
16. Nakao, S.K.; Fassler, S.; Sucandy, I.; Kim, S.; Zebley, D.M. Colorectal cancer following negative colonoscopy: Is 5-year screening
the correct interval to recommend? Surg. Endosc. 2012, 27, 768–773. [CrossRef] [PubMed]
17. Pai, R.K.; Bettington, M.; Srivastava, A.; Rosty, C. An update on the morphology and molecular pathology of serrated colorectal
polyps and associated carcinomas. Mod. Pathol. 2019, 32, 1390–1415. [CrossRef]
18. Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature 2012, 487,
330–337. [CrossRef] [PubMed]
19. Giannakis, M.; Mu, X.J.; Shukla, S.A.; Qian, Z.R.; Cohen, O.; Nishihara, R.; Bahl, S.; Cao, Y.; Amin-Mansour, A.; Yamauchi, M.;
et al. Genomic correlates of immune-cell infiltrates in colorectal carcinoma. Cell Rep. 2016, 15, 857–865. [CrossRef] [PubMed]
20. Balkwill, F.; Mantovani, A. Inflammation and cancer: Back to virchow? Lancet 2001, 357, 539–545. [CrossRef]
21. Lai, H.-W.; Loong, C.-C.; Tai, L.-C.; Wu, C.-W.; Lui, W.-Y. Incidence and odds ratio of appendicitis as first manifestation of colon
cancer: A retrospective analysis of 1873 patients. J. Gastroenterol. Hepatol. 2006, 21, 1693–1696. [CrossRef]
22. Yuan-Kun, C.; Tao, Y. Does a decrease of NK cells in the appendix increase the risk of developing colon cancer? Hepato-
Gastroenterology 2012, 59, 1819–1821. [PubMed]
23. Feagins, L.A.; Souza, R.F.; Spechler, S.J. Carcinogenesis in IBD: Potential targets for the prevention of colorectal cancer. Nat. Rev.
Gastroenterol. Hepatol. 2009, 6, 297–305. [CrossRef] [PubMed]
24. Rubin, D.C.; Shaker, A.; Levin, M.S. Chronic intestinal inflammation: Inflammatory bowel disease and colitis-associated colon
cancer. Front. Immunol. 2012, 3, 107. [CrossRef] [PubMed]
25. Ullman, T.A.; Itzkowitz, S.H. Intestinal inflammation and cancer. Gastroenterology 2011, 140, 1807–1816.e1. [CrossRef]
26. Grivennikov, S.I. Inflammation and colorectal cancer: Colitis-associated neoplasia. Semin. Immunopathol. 2013, 35, 229–244.
[CrossRef] [PubMed]
27. Grivennikov, S.I.; Wang, K.; Mucida, D.; Stewart, C.A.; Schnabl, B.; Jauch, D.; Taniguchi, K.; Yu, G.Y.; Osterreicher, C.H.; Hung,
K.E.; et al. Adenoma-linked barrier defects and microbial products drive IL-23/IL-17-mediated tumour growth. Nature 2012, 491,
254–258. [CrossRef]
28. Chapelle, N.; Martel, M.; Toes-Zoutendijk, E.; Barkun, A.N.; Bardou, M. Recent advances in clinical practice: Colorectal cancer
chemoprevention in the average-risk population. Gut 2020, 69, 2244–2255. [CrossRef]
29. Drew, D.A.; Chin, S.; Gilpin, K.K.; Parziale, M.; Pond, E.; Schuck, M.M.; Stewart, K.; Flagg, M.; Rawlings, C.A.; Backman, V.; et al.
ASPirin intervention for the REDuction of colorectal cancer risk (ASPIRED): A study protocol for a randomized controlled trial.
Trials 2017, 18, 50. [CrossRef]
30. Tseng, H.-F.; Morgenstern, H.; Mack, T.M.; Peters, R.K. Risk factors for anal cancer: Results of a population-based case—Control
study. Cancer Causes Control. 2003, 14, 837–846. [CrossRef] [PubMed]
31. Holly, E.A.; Whittemore, A.S.; Aston, D.A.; Ahn, D.K.; Nickoloff, B.J.; Kristiansen, J.J. Anal cancer incidence: Genital warts, anal
fissure or fistula, hemorrhoids, and smoking. J. Natl. Cancer Inst. 1989, 81, 1726–1731. [CrossRef]
Int. J. Environ. Res. Public Health 2021, 18, 8655 10 of 10

32. Ungerbäck, J.; Belenki, D.; Ul-Hassan, A.J.; Fredrikson, M.; Fransén, K.; Elander, N.; Verma, D.; Söderkvist, P. Genetic variation
and alterations of genes involved in NFκB/TNFAIP3- and NLRP3-inflammasome signaling affect susceptibility and outcome of
colorectal cancer. Carcinogenesis 2012, 33, 2126–2134. [CrossRef] [PubMed]
33. Yuhara, H.; Steinmaus, C.; Cohen, S.E.; Corley, D.A.; Tei, Y.; Buffler, P.A. Is diabetes mellitus an independent risk factor for colon
cancer and rectal cancer? Am. J. Gastroenterol. 2011, 106, 1911–1921. [CrossRef] [PubMed]
34. Lee, P.-C.; Hu, Y.-W.; Hung, M.-H.; Chen, C.-C.; Lin, H.-C.; Lee, F.-Y.; Hung, Y.-P.; Su, V.Y.-F.; Yen, S.-H.; Tzeng, C.-H.; et al. The
risk of cancer in patients with benign anal lesions: A nationwide population-based study. Am. J. Med. 2013, 126, 1143.e9–1143.e18.
[CrossRef] [PubMed]