Open Journal of Gastroenterology, 2013, 3, 25-34 OJGas
http://dx.doi.org/10.4236/ojgas.2013.31004 Published Online February 2013 (http://www.scirp.org/journal/ojgas/)
Risk factors for internal anal sphincter dysfunction in
Japanese adults
Tatsuya Abe1*, Toru Kono2,3, Yoshikazu Hachiro1, Masao Kunimoto1, Hiroyuki Furukawa2
1Department of Proctology, Kunimoto Hospital, Asahikawa, Japan
2Division of Gastroenterologic and General Surgery, Department of Surgery, Asahikawa Medical University, Asahikawa, Japan
3Advanced Surgery Center, Sapporo Higashi Tokushukai Hospital, Sapporo, Japan
Received 13 December 2012; revised 13 January 2013; accepted 20 January 2013
ABSTRACT
Purpose: The internal anal sphincter provides most of
the resting anal tone and is the main muscle response-
ble for continence. This study was designed to esti-
mate the prevalence of, and identify risk factors asso-
ciated with, internal anal sphincter dysfunction in
Japanese adults. Methods: Anorectal manometry was
performed in 1193 women and 1124 men aged 20
years or older. The maximal resting pressure, meas-
ured by a rapid pull-through technique, was defined
as the highest resting pressure recorded. Internal
anal sphincter dysfunction was defined as a maximal
resting pressure less than 30 mmHg. Potential risk
factors were assessed through self-reports, interviews,
physical examinations, and medical record reviews.
Multivariate logistic regression analysis was used to
identify independent risk factors for internal anal
sphincter dysfunction. Results: Significant differences
in maximal resting pressure wer e seen bet ween w omen
(58.1 ± 24.9 mmHg) and men (68.8 ± 23.5 mmHg, P <
0.001). Maximal resting pressure decreased signifi-
cantly with increasing age in both sexes. The preva-
lence of internal anal sphincter dysfunction was 1 0 .4%
(15.5% in women, 5.1% in men). In a multivariate
logistic regression model, age, mental disease, pelvic
organ prolapse repair, and fecal incontinence were
independently associated with a greater risk of inter-
nal anal sphincter dysfunction in women and men.
Conclusions: Internal anal sphincter dysfunction is a
common problem for women and men. Several of the
identified risk factors are preventable or modifiable,
and may direct future research in fecal incontinence
therapy.
Keywords: Fecal Incontinence; Internal Anal Sphincter;
Anorectal Manometry; Maximal Resting Pressure
1. INTRODUCTION
Fecal incontinence (FI) is a common disorder with a sig-
nificant impact on quality of life. The prevalence of FI in
the general population increases with age in both women
and men [1]. Although the cause of FI is often multifac-
torial, FI is frequently caused by anal sphincter insuffi-
ciency [2]. The anal sphincter consist of the circular in-
ternal (IAS) and external (EAS) anal sphincter muscles
together with the sling-shaped puborectalis muscle. Al-
though both sphincter muscles are important for the main-
tenance of continence, the IAS, composed of smooth
muscle arranged in oblique bundles, provides most of the
resting anal tone and is the main muscle responsible for
preventing fecal leakage. The IAS contributes an esti-
mated 55% - 85% to maximal resting pressure (MRP)
[3,4]. Low MRP is the most important predictor of FI
and correlates with the Fecal Incontinence Severity Index
(FISI) [5].
Despite a number of studies on FI in western countries
[6-8], there has been a paucity of research about risk
factors associated with IAS dysfunction (IASD). Identi-
fying preventable or modifiable risk factors for IASD
may guide future research for the prevention or treatment
of FI.
The objective of this study was to provide a compre-
hensive description of IASD in Japanese adults and to
describe demographic and other risk factors associated
with IASD after multivariate adjustments.
2. MATERIALS AND METHODS
Between January 2006 and December 2008, 3190 con-
secutive subjects aged 20 years or older referred to our
hospital were enrolled.
Factors potentially associated with IASD or FI were
assessed by questionnaires, interviews, medical record
reviews, and physical examinations. Participants filled in
a structured questionnaire, which included questions
about demographic characteristics (age, sex, body mass
*Corresponding author.

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 26
index (BMI), employment), stool consistency, bowel
movements, presence of selected medical conditions,
previous anorectal surgeries, previous cholecystectomy,
current habits, laxative use, FI, and parity in women. All
medical conditions and medication use were determined
according to self-reports as well as in-person interviews
during which written responses were clarified. Patient
weight and height were measured, and BMI was catego-
rized according to the classification proposed for Japa-
nese adults. The validated Bristol Stool Scale was used to
describe the participant’s usual stool consistency [9].
This scale consists of 7 types of stool and includes pic-
tures of each stool type to aid participants in classifying
their stools. We combined stool type 1 and 2 (hard and
lumpy), 3 - 5 (normal consistencies), and 6 and 7 (mushy
and watery). Smoking status was categorized as non-
smoker and current smoker (≥1 cigarettes per day). Al-
cohol intake was categorized as nondrinker and current
drinker (more than once a week). For this study, FI was
defined as any involuntary loss of mucus, liquid, or solid
stool during the last 30 days; this definition of FI did not
include gas.
Anorectal manometry (ARM) was performed by a sin-
gle experienced examiner using a 5 mm diameter, 1-
channel solid-state catheter with a microtipped trans-
ducer ARM system (P-31, Star Medical Co., Tokyo, Ja-
pan). All subjects were examined in the left lateral posi-
tion with the hips flexed to 90˚. The lubricated catheter
was introduced into the rectum. The MRP measured by
means of a rapid pull-through technique was defined as
the highest resting pressure recorded. IASD was defined
as “low MRP” in cases where MRP was less than 30
mmHg [10].
Statistical analysis was performed using JMP version
9.0.2 (SAS Institute Inc, Cary, NC). MRP values were
expressed as means ± standard deviation. The Mann-
Whitney U test was used to analyze for sex differences in
MRP, and regression analysis was used to search for as-
sociations with age. The chi-square test was used to test
the association between low MRP and each risk factor
individually. All risk factors found to be significantly
associated with low MRP in univariate analyses were
combined into separate multivariate logistic models for
women and men. Adjusted odds ratios (OR) and 95%
confidence intervals (CI) were used to describe associa-
tions between risk factors and low MRP in univariate
analyses and in multivariate analyses. P values < 0.05
were considered to indicate statistical significance for all
analyses.
This study was a retrospective review of existing
clinical data prospectively collected on a hospital anal
physiology unit computer database. This study was ap-
proved by the research and ethics committee of Kuni-
moto Hospital and informed consent for ARM and future
use of their data was provided by all participants in the
study.
3. RESULTS
Data from 2317 subjects (1193 women and 1124 men)
who gave complete responses were analyzed. Mean age
was 56.0 ± 17.9 years in women and 53.2 ± 16.1 years in
men. All participants were stratified into 10-year age
groups. The general characteristics of the participants are
shown in Ta bl e 1 . The most common medical condition
was hypertension (26.8%). FI was reported by 15.1% of
women and 7.7% of men.
Significant differences in MRP were seen between
women (58.1 ± 24.9 mmHg) and men (68.8 ± 23.5 mmHg;
P < 0.001). MRP decreased significantly with increasing
age in both sexes (Figure 1).
The prevalence of low MRP was 10.4% (15.5% in
women, 5.1% in men). Among women, prevalence in-
creased from 1.7% for 20 - 29 year-olds up to 52.4% for
80 years or older. For women and men combined, the
most common medical conditions associated with low
MRP were cognitive impairment and pelvic organ prolapse
repair (Tables 2 and 3).
In univariate analyses, variables associated with low
MRP in women and men were age, employment status,
hypertension, stroke, cognitive impairment, mental dis-
ease, ischemic heart disease, any history of cancer, pelvic
organ prolapse repair, laxative use, and FI. In addition,
parity, dyslipidemia, spinal cord injury, anal fistula sur-
gery, and hysterectomy were associated with low MRP in
women but not in men, while BMI, usual stool consis-
tency, diabetes mellitus, hemorrhoidectomy, and rectal
surgery were associated with low MRP in men alone.
In a multivariate logistic regression model, age, men
tal disease, pelvic organ prolapse repair, and FI were
independently associated with a greater risk of low MRP
in women and men. In addition, ischemic heart disease
was associated with low MRP in women alone. By con-
trast, smoking or alcohol intake and diabetes mellitus
were significantly negative associated with low MRP in
women (Tables 4 and 5).
4. DISCUSSION
The present study investigated the influence of age, sex,
and other possible risk factors for low MRP in a large
number of Japanese adults spanning a wide age range.
The main results are the following: 1) women had sig-
nificantly lower MRP than men; 2) MRP decreased with
increasing age in both sexes; 3) independent risk factors
for low MRP in both sexes include advancing age, FI,
mental disease, and pelvic organ prolapse repair.
FI is defined as either the involuntary passage of or the
inability to control the discharge of fecal matter through

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 27
Table 1. General characteristics of study subjects.
Characteristic Overall
n = 2317 Women n = 1193 Men n = 1124
Age (year) 20 - 29 201 (8.7) 117 (9.8) 84 (7.5)
30 - 39 348 (15.0) 149 (12.5) 199 (17.7)
40 - 49 339 (14.6) 157 (13.2) 182 (16.2)
50 - 59 470 (20.3) 216 (18.1) 254 (22.6)
60 - 69 421 (18.2) 231 (19.4) 190 (16.9)
70 - 79 391 (16.9) 220 (18.4) 171 (15.2)
≥80 147 (6.3) 103 (8.6) 44 (3.9)
BMI Underweight (<18.5) 140 (6.0) 87 (7.3) 53 (4.7)
Normal (18.5 - 24.9) 1628 (70.3) 898 (75.3) 730 (64.9)
Overweight (≥25) 549 (23.7) 208 (17.4) 341 (30.3)
Parity At least one vaginal 924 (77.4)
Employment status Employed/student 1281 (55.3) 470 (39.4) 811 (72.2)
Unemployed/other 1036 (44.7) 723 (60.6) 313 (27.8)
Usual stool consistency Normal stools 1403 (60.6) 635 (53.2) 768 (68.3)
Hard, lumpy stools 688 (29.7) 466 (39.1) 222 (19.8)
Loose, watery stools 150 (6.5) 47 (3.9) 103 (9.2)
Other 76 (3.3) 45 (3.8) 31 (2.8)
Frequency of bowel movements >21 BM/week 113 (4.9) 49 (4.1)
64 (5.7)
3 - 21 BM/week 2001 (86.4) 1001 (83.9) 1000 (89.0)
<3 BM/week 203 (8.8) 143 (12.0) 60 (5.3)
Comorbid conditions Hypertension 620 (26.8) 317 (26.6) 303 (27.0)
Diabetes mellitus 155 (6.7) 75 (6.3) 80 (7.1)
Dyslipidemia 257 (11.1) 154 (12.9) 103 (9.2)
Stroke 93 (4.0) 44 (3.7) 49 (4.4)
Cognitive impairment 29 (1.3) 17 (1.4) 12 (1.1)
Mental disease 102 (4.4) 60 (5.0) 42 (3.7)
Ischemic heart disease 64 (2.9) 27 (2.3) 37 (3.3)
Any history of cancer 108 (4.7) 58 (4.9) 50 (4.4)
Bronchial asthma 98 (4.2) 57 (4.8) 41 (3.6)
Spinal cord injury 46 (2.0) 22 (1.8) 24 (2.1)
Previous surgeries Hemorrhoidectomy 314 (13.6) 156 (13.1) 158 (14.1)
Anal fissure surgery 28 (1.2) 16 (1.3) 12 (1.1)
Anal fistula surgery 82 (3.5) 16 (1.3) 66 (5.9)
Hysterectomy 97 (8.1)
Pelvic organ prolapse repair 51 (2.2) 43 (3.6) 8 (0.7)
Rectal surgery 16 (0.7) 7 (0.6) 9 (0.8)
Cholecystectomy 43 (1.9) 17 (1.4) 26 (2.3)
Current habits Smoking 572 (24.7) 194 (16.3) 378 (33.6)
Alcohol intake 979 (42.3) 293 (24.6) 686 (61.0)
Laxative use 471 (20.3) 347 (29.1) 124 (11.0)
Fecal incontinence 266 (11.5) 180 (15.1) 86 (7.7)
BMI = body mass index. BM = bowel movements. Data are numbers of subjects with percentages in parentheses. Percentages may not sum to 100 because of
rounding.

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 28
(a) (b)
Figure 1. Regression analysis on the influence of age on maximal resting pressure (MRP) in women (a) and men (b). MRP decreased
significantly with increasing age in both sexes (women, r = –0.721, P < 0.001; men, r = –0.583, P < 0.001).
the anal canal [1]. Clinically there are 3 subtypes: 1) pas-
sive incontinence—the involuntary discharge of rectal
contents without awareness; 2) urge incontinence—the
discharge of stool in spite of active attempts to retain
bowel contents; and 3) combined incontinence—both
passive and urge incontinence [2]. The majority of FI
patients have passive incontinence [11]. Passive inconti-
nence is generally associated with dysfunction of the
smooth muscle of the IAS, whereas urge incontinence is
related to dysfunction of the striated muscle of the EAS
[1,2]. At rest, the IAS is in a tonically contracted state,
and is innervated by the autonomic nervous system. The
IAS plays a key role in maintaining continence, provid-
ing 55% - 85% of MRP [3,4].
Several studies have investigated the effect of sex on
resting pressure of the anal canal [12-15]. The data are
contradictory, with about half of the studies reporting
comparable resting pressures in women and men, and the
other half citing lower resting pressure in women. In
most of these studies, the number of male and female
subjects was rather low. Laurberg and Swash [12], who
studied a larger cohort of subjects (102 women and 19
men), observed a significantly lower resting pressure in
women. This is in accordance with the results of our
study on more than 1000 people of both sexes.
Reports on the effects of aging on MRP are also con-
flicting. Some studies observed a significant lowering of
MRP with age (more often in women than in men), while
others found only a small decrease in MRP that did not
reach statistical significance. Other studies did not see a
decreased MRP in their older subjects [12-15]. In addi-
tion to the aforementioned problem of small sample size,
most studies included patients of varying ages and grouped
them differently. In the present study, we divided the
subjects into 7 age groups by decade, and used regres-
sion analysis to search for effects of age on MRP, and
found that MRP significantly decreases with age in
women as well as in men. This age-dependent decrease
in MRP is thought to be due to fibrosis of the IAS in eld-
erly patients [16,17].
Studies examining independent risk factors for FI have
implicated age, obstetric and gynecologic factors, several
medical conditions, and poor health. In particular, age,
parity, vaginal delivery, diabetes, diarrhea, and neurologic
conditions have notably been associated [6-8,18,19].
This study has several limitations that should be con-
sidered when interpreting the results. First, prevalence
estimates were limited to non-institutionalized individu-
als, and because the prevalence of FI in the nursing home
population is higher than in the community [1,2], the
overall prevalence of IASD is likely underestimated,
especially among older participants. Second, as in pre-
vious large epidemiological studies, comorbid conditions
were determined according to self-reports instead of
through clinician-derived methods.
Since women were found to have a higher prevalence
of FI than men, in some prior studies, it was hypothe-
sized that this sex difference is due to obstetrical injuries
being a major risk factor for FI [18,19]. In this survey of
the entire adult lifespan, there was a significant differ-
ence in the prevalence of low MRP between women and
men. In univariate analysis, vaginal delivery was associ-
ated with a higher prevalence of low MRP in women, but
this association was no longer significant after multivari-
ate adjustment. Whitehead et al. [7] did not find a sig-
nificant difference in the prevalence of FI between
women and men comprising the entire adult lifespan.
Other studies that included a broad range of ages have

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 29
Table 2. Baseline characteristics of the risks for low MRP of women.
Risk factor Women
n = 1193 Low MRP n = 185 (15.5) P
Age (year) 20 - 29 117 (9.8) 2 (1.7) <0.001
30 - 39 149 (12.5) 4 (2.7)
40 - 49 157 (13.2) 5 (3.1)
50 - 59 216 (18.1) 9 (4.5)
60 - 69 231 (19.4) 42 (18.2)
70 - 79 220 (18.4) 69 (31.4)
≥80 103 (8.6) 54 (52.4)
BMI Underweight (<18.5) 87 (7.3) 20 (23.0) 0.13
Normal (18.5 - 24.9) 898 (75.3) 135 (15.0)
Overweight (≥25) 208 (17.4) 30 (14.4)
Parity At least one vaginal 924 (77.4) 173 (18.7) <0.001
None 269 (22.6) 12 (4.5)
Employment status Employed/student 470 (39.4) 33 (7.0) <0.001
Unemployed/other 723 (60.6) 152 (21.0)
Usual stool consistency Normal stools 635 (53.2) 102 (16.1) 0.41
Hard, lumpy stools 466 (39.1) 73 (15.7)
Loose, watery stools 47 (3.9) 7 (14.9)
Other 45 (3.8) 3 (6.7)
Frequency of bowel movements >21 BM/week 49 (4.1) 12 (24.5) 0.06
3 - 21 BM/week 1001 (83.9) 145 (14.5)
<3 BM/week 143 (12.0) 28 (19.6)
Comorbid conditions Hypertension 317 (26.6) 81 (25.6) <0.001
Diabetes mellitus 75 (6.3) 16 (21.3) 0.15
Dyslipidemia 154 (12.9) 34 (22.1) 0.01
Stroke 44 (3.7) 13 (29.6) 0.009
Cognitive impairment 17 (1.4) 10 (58.8) <0.001
Mental disease 60 (5.0) 17 (28.3) 0.005
Ischemic heart disease 27 (2.3) 15 (55.6) <0.001
Any history of cancer 58 (4.9) 16 (27.6) 0.009
Bronchial asthma 57 (4.8) 13 (22.8) 0.12
Spinal cord injury 22 (1.8) 11 (50.0) <0.001
Previous surgeries Hemorrhoidectomy 156 (13.1) 29 (18.6) 0.25
Anal fissure surgery 16 (1.3) 4 (25.0) 0.29
Anal fistula surgery 16 (1.3) 6 (37.5) 0.01
Hysterectomy 97 (8.1) 25 (25.8) 0.004
Pelvic organ prolapse repair 43 (3.6) 25 (58.1) <0.001
Rectal surgery 7 (0.6) 3 (42.9) 0.08
Cholecystectomy 17 (1.4) 5 (29.4) 0.11
Current habits Smoking 194 (16.3) 5 (2.6) <0.001
Alcohol intake 293 (24.6) 7 (2.4) <0.001
Laxative use 347 (29.1) 77 (22.2) <0.001
Fecal incontinence 180 (15.1) 103 (57.2) <0.001
MRP = maximal resting pressure. BMI = body mass index. BM = bowel movements. Data are numbers of subjects with percentages in parentheses. Percentages
may not sum to 100 because of rounding.

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 30
Table 3. Baseline characteristics of the risks for low MRP of men.
Risk factor Men
n = 1124 Low MRP n = 57 (5.1) P
Age (year) 20 - 29 84 (7.5) 1 (1.2) <0.001
30 - 39 199 (17.7) 2 (1.0)
40 - 49 182 (16.2) 1 (0.6)
50 - 59 254 (22.6) 3 (1.2)
60 - 69 190 (16.9) 11 (5.8)
70 - 79 171 (15.2) 26 (15.2)
≥80 44 (3.9) 13 (29.5)
BMI Underweight (<18.5) 53 (4.7) 5 (9.4) 0.01
Normal (18.5 - 24.9) 730 (64.9) 44 (6.0)
Overweight (≥25) 341 (30.3) 8 (2.3)
Employment status Employed/student 811 (72.2) 17 (2.1) <0.001
Unemployed/other 313 (27.8) 40 (12.8)
Usual stool consistency Normal stools 768 (68.3) 27 (3.5) 0.005
Hard, lumpy stools 222 (19.8) 20 (9.0)
Loose, watery stools 103 (9.2) 7 (6.8)
Other 31 (2.8) 3 (9.7)
Frequency of bowel movements >21 BM/week 64 (5.7) 6 (9.4) 0.27
3 - 21 BM/week 1000 (89.0) 48 (4.8)
<3 BM/week 60 (5.3) 3 (5.0)
Comorbid conditions Hypertension 303 (27.0) 25 (8.6) 0.003
Diabetes
mellitus 80 (7.1) 8 (10.0) 0.04
Dyslipidemia 103 (9.2) 6 (5.8) 0.71
Stroke 49 (4.4) 6 (12.2) 0.02
Cognitive impairment 12 (1.1) 7 (58.3) <0.001
Mental disease 42 (3.7) 8 (19.1) <0.001
Ischemic heart disease 37 (3.3) 5 (13.5) 0.02
Any history of cancer 50 (4.4) 6 (12.0) 0.02
Bronchial asthma 41 (3.6) 3 (7.3) 0.50
Spinal cord injury 24 (2.0) 2 (8.3) 0.46
Previous surgeries Hemorrhoidectomy 158 (14.1) 19 (12.0) <0.001
Anal fissure surgery 12 (1.1) 1 (8.3) 0.61
Anal fistula surgery 66 (5.9) 6 (9.1) 0.13
Pelvic organ prolapse repair 8 (0.7) 4 (50.0) <0.001
Rectal surgery 9 (0.8) 3 (33.3) <0.001
Cholecystectomy 26 (2.3) 1 (3.8) 0.77
Current habits Smoking 378 (33.6) 15 (4.0) 0.37
Alcohol intake 686 (61.0) 25 (3.6) 0.007
Laxative use 124 (11.0) 19 (15.3) <0.001
Fecal incontinence 86 (7.7) 26 (30.2) <0.001
MRP = maximal resting pressure. BMI = body mass index. BM = bowel movements. Data are numbers of subjects with percentages in parentheses. Percentages
may not sum to 100 because of rounding.

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 31
Table 4. Odds ratios for variables associated with low MRP of women.
Risk factor Univariate analysis
OR (95% CI) P Multivariate analysis OR (95% CI) P
Age (10-year interval) 2.34 (2.04, 2.71) <0.0011.74 (1.44, 2.12) <0.001
BMI (vs. normal)
Underweight 1.69 (0.97, 2.82) 0.06 2.05 (0.92, 4.42) 0.08
Overweight 0.95 (0.61, 1.44) 0.82 0.70 (0.41, 1.16) 0.17
Parity (vs. none) 4.93 (2.82, 9.49) <0.0011.83 (0.87, 4.17) 0.11
Employment status (vs. employed/student)
Unemployed/other 3.53 (2.40, 5.32) <0.0011.49 (0.86, 2.56) 0.15
Usual stool consistency (vs. normal stools)
Hard, lumpy stools 0.97 (0.70, 1.34) 0.86 0.75 (0.46, 1,23) 0.26
Loose, watery stools 0.91 (0.37, 1.98) 0.83 0.93 (0.33, 2.83) 0.89
Other 0.37 (0.09, 1.05) 0.06 0.81 (0.17, 2.92) 0.77
Comorbid condition
Hypertension 2.55 (1.84, 3.53) <0.0010.74 (0.46, 1.16) 0.19
Diabetes mellitus 1.52 (0.85, 2.71) 0.15 0.40 (0.18, 0.82) 0.01
Dyslipidemia 1.67 (1.10, 2.53) 0.01 1.41 (0.82, 2.38) 0.21
Stroke 2.38 (1.22, 4.64) 0.009 0.47 (0.19, 1.12) 0.09
Cognitive impairment 8.17 (3.07, 21.8) <0.0010.83 (0.24, 3.07) 0.78
Mental disease 2.27 (1.27, 4.08) 0.005 2.23 (1.00, 4.84) 0.049
Ischemic heart disease 7.32 (3.37, 15.9) <0.0013.14 (1.23, 8.22) 0.02
Any history of cancer 2.18 (1.20, 3.96) 0.009 1.11 (0.47, 2.50) 0.81
Spinal cord injury 5.73 (2.45, 13.4) <0.0012.51 (0.82, 7.58) 0.10
Previous surgeries
Hemorrhoidectomy 1.29 (0.83, 2.00) 0.25 0.97 (0.55, 1.65) 0.91
Anal fistula surgery 3.35 (1.20, 9.32) 0.01 1.95 (0.48, 7.04) 0.33
Hysterectomy 2.03 (1.25, 3.30) 0.004 1.57 (0.82, 2.95) 0.17
Pelvic organ prolapse repair 8.59 (4.58, 16.1) <0.0012.48 (1.07, 5.84) 0.03
Rectal surgery 4.14 (0.80, 18.9) 0.08 1.11 (0.17, 6.75) 0.91
Current habits
Smoking 0.12 (0.04, 0.27) <0.0010.25 (0.10, 0.56) <0.001
Alcohol intake 0.10 (0.04, 0.20) <0.0010.27 (0.08, 0.72) 0.007
Laxative use 1.95 (1.41, 2.69) <0.0011.29 (0.81, 2.05) 0.29
Fecal incontinence 15.17 (10.5, 22.1) <0.0018.16 (5.20, 12.9) <0.001
also failed to find a sex difference in FI prevalence or an
effect of obstetrical injury [6,8,20]. This suggests that
obstetrical injuries are not the most common cause of FI
in women. Perineal tears are classified into 4 degrees. A
fourth-degree tear is defined as an injury that extends
through the fascia and musculature of the perineal body
and involves the anal sphincter complex (EAS and IAS)
and anal epithelium. Third-degree tears involve some or
all of the fibers of the EAS, but do not necessarily in-
volve the IAS. Faltin et al. [21] reported that clinically
undetected tears of the anal sphincter were diagnosed by
anal endosonography in 42 of 150 women (28%). EAS
injury alone was observed in 30 women (20%), IAS in-
jury alone in 2 (1.3%), and both in 10 (7%). In a study of
62 women with FI related to obstetrical procedures, anal
endosonography revealed EAS defects in 90% and IAS
defects in 65% of the subjects [22]. Our data also sug-
gests that EAS injury is the predominant cause of FI after
vaginal delivery.
Previous studies have found diabetes mellitus to in-
crease the risk of FI [8,23]. Ward and Tunuguntla [23]
reported that 4% of diabetic patients suffered from FI.

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 32
Table 5. Odds ratios for variables associated with low MRP of men.
Risk factor Univariate analysis
OR (95% CI) P Multivariate analysis OR (95% CI) P
Age (10-year interval) 2.54 (2.02, 3.29) <0.0012.36 (1.68, 3.41) <0.001
BMI (vs. Normal)
Underweight 1.62 (0.54, 3.94) 0.35 1.15 (0.36, 4.55) 0.82
Overweight 0.37 (0.16, 0.76) 0.006 0.49 (0.19, 1.13) 0.10
Employment status (vs. Employed/student)
Unemployed/Other 6.84 (3.88, 12.6) <0.0011.10 (0.48, 2.49) 0.81
Usual stool consistency (vs. normal stools)
Hard, lumpy stools 2.72 (1.48, 4.93) 0.002 1.94 (0.59, 5.68) 0.27
Loose, watery stools 2.00 (0.79, 4.48) 0.14 1.57 (0.68, 3.52) 0.29
Other 2.94 (0.67, 8.99) 0.13 2.30 (0.37, 10.3) 0.85
Comorbid condition
Hypertension 2.22 (1.29, 3.81) 0.003 0.67 (0.31, 1.41) 0.30
Diabetes mellitus 2.26 (1.03, 4.95) 0.04 0.97 (0.33, 2.53) 0.96
Dyslipidemia 1.18 (0.49, 2.81) 0.71 0.86 (0.27, 2.29) 0.78
Stroke 2.80 (1.14, 6.89) 0.02 0.48 (0.13, 1.47) 0.21
Cognitive impairment 29.73 (9.11, 98.0) <0.0014.37 (0.97, 21.5) 0.06
Mental disease 4.96 (2.18, 11.3) <0.0017.29 (2.36, 21.2) <0.001
Ischemic heart disease 3.11 (1.16, 8.31) 0.02 1.98 (0.53, 6.32) 0.29
Any history of cancer 2.74 (1.11, 6.72) 0.02 0.62 (0.16, 1.98) 0.44
Spinal cord injury 1.73 (0.40, 7.53) 0.46 1.40 (0.19, 5.98) 0.70
Previous surgeries
Hemorrhoidectomy 3.34 (1.87, 5.96) <0.0011.55 (0.72, 3.21) 0.26
Anal fistula surgery 1.97 (0.81, 4.78) 0.13 2.02 (0.57, 6.02) 0.26
Pelvic organ prolapse repair 20.06 (4.88, 82.4) <0.00111.87 (1.71, 88.9) 0.01
Rectal surgery 9.82 (2.39, 40.3) <0.0015.94 (0.76, 43.9) 0.09
Current habits
Smoking 0.69 (0.37, 1.24) 0.37 0.79 (0.65, 2.42) 0.50
Alcohol intake 0.47 (0.28, 0.82) 0.007 1.15 (0.53, 2.40) 0.72
Laxative use 4.58 (2.55, 8.23) <0.0011.57 (0.68, 3.51) 0.28
Fecal incontinence 14.08 (7.83, 25.23) <0.0015.13 (2.14, 12.12) 0.001
MRP = maximal resting pressure. OR = odds ratio. CI = confidence interval. BMI = body mass index. Only variables found to be significant in univariate
analyses for women or men were included in the multivariate regressions and shown in the table. Multivariate odds ratios are adjusted for all other risk factors
in the table.
The underlying pathophysiology of anorectal dysfunction
in diabetic patients with FI is not well understood. Dia-
betes mellitus may contribute to FI through sphincter or
pelvic floor weakness from anatomic defects, nerve dam-
age, or microvascular complications. Schiller et al. [24]
found IASD secondary to autonomic neuropathy respon-
sible for FI in 16 diabetic patients. In contrast, Rogers
[25] reported that EAS dysfunction might play an im-
portant role. Other reports suggested that both IAS and
EAS dysfunction might be involved [26,27]. In the pre-
sent study, the univariate association between diabetes
mellitus and low MRP was significant for men, but after
multivariate adjustment, this relationship did not hold.
On the contrary, in women low MRP was significantly
negatively associated with diabetes mellitus after multi-
variate adjustment. The lumbar sympathetic innervation
inhibits the colon and provides motor stimulation to the
IAS. The pudendal nerves, on the other hand, are somatic
nerves that innervate the EAS. Diabetic pudendal neu-
ropathy leads to delay of pudendal nerve terminal motor
latency, causes EAS damage, and may result in FI [28].
Our results suggest that somatic neuropathy plays a more
important role in FI in diabetic patients.
FI was significantly associated with low MRP after

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 33
multivariate adjustment in both sexes. More treatment
options for FI makes a proper pretreatment evaluation
increasingly important, whether by means of subjective
severity questionnaires, such as the FISI, or by objective
measurements, such as ARM, anal endosonography,
electromyography, and defecography [1]. Bordeianou et
al. [5] reported MRP was the only objective measure-
ment that seems to correlate with both FISI and the
presence of sphincteric defects on anal endosonography,
while maximal squeeze pressure (MSP) generated by the
EAS was not correlated. Our results confirm a strong
association between low MRP and FI.
Mental disease was found to be independently associ-
ated with low MRP. The IAS has spontaneous myogenic
tone, but it also receives inputs from the sympathetic and
parasympathetic nervous systems, both of which exert
excitatory and inhibitory influences on the sphincteric
smooth muscle [29]. Yamato and Rattan [30] demon-
strated that stimulation of α1-adrenoreceptors causes an
increase in the resting pressure developed by the IAS.
Multi-acting receptor targeted antipsychotics (MARTA)
have a high affinity for a number of receptors including
the dopaminergic D2, serotonergic 5-HT2A, α1-adren-
ergic, and muscarinic receptors. Two cases of FI have
been reported in patients taking MARTA [31,32]. The
pathophysiology is thought to be related to its central
mechanism of action and inhibitory effects on the anal
sphincters. Siproudhis et al. [33] reported that a single
oral administration of 2 types of antidepressants (amitrip-
tyline and fluoxetine) reduced pressure in the upper anal
canal in 10 healthy male volunteers. Both drugs are
likely to decrease the contractile activity of smooth mus-
cle by inhibiting calcium channels.
Pelvic organ prolapse repair was also an independent
risk factor for low MRP. Most studies reported a greater
than 10% prevalence of FI in women with genital
prolapse, and more than a third of women and men with
rectal prolapse reported FI [34]. Internal rectal prolapse
is a circular infolding of the rectal wall that occurs on
straining to defecate. It descends into the rectum and
may reach into the anal canal. Harmston et al. [35] found
a significant reduction in MRP with progressive in-
creases in the grade of internal rectal prolapse without a
demonstrable effect on MSP. By contrast, external rectal
prolapse was associated with a significant reduction in
both MRP and MSP [36]. The underlying mechanism has
been postulated to result from nerve damage by stretch-
ing due to perineal descent or sphincter dilatation and
inappropriate stimulation of the recto-anal inhibitory re-
flex by the prolapsing rectum [37]. The finding of de-
creased MRP without a significant change in MSP sug-
gests that the predominant effect of internal rectal prolapse
is on reducing IAS tone without an effect on EAS.
5. CONCLUSION
IASD has received little attention, and the results of this
study indicate that it is a major problem for elderly peo-
ple in Japan. In addition, the importance of IASD as a
public health problem is likely to increase in the near
future as the elderly population continues to grow. This
cross-sectional study offers evidence that IASD is corre-
lated with the presence of certain conditions such as age,
sex, FI, mental disease, and pelvic organ prolapse. Sev-
eral of these risk factors for IASD are potentially pre-
ventable or modifiable; for example, several studies have
shown a recovery in MRP after rectal prolapse repair
[35]. Improving our understanding of risk factors, par-
ticularly modifiable risk factors, is critical for developing
future prevention guidelines and improving the specific-
ity of FI treatment.
REFERENCES
[1] Madoff, R.D., Parker, S.C., Varma, M.G., et al. (2004)
Faecal incontinence in adults. Lancet, 364, 621-632.
doi:10.1016/S0140-6736(04)16856-6
[2] Rao, S.S.C. (2004) Diagnosis and management of fecal
incontinence. American Journal of Gastroenterology, 99,
1585-1604. doi:10. 1111/j.1572-0241.2004.40105.x
[3] Frenckner, B. and Euler, C.V. (1975) Influence of pu-
dendal block on the function of the anal sphincters. Gut,
16, 482-489. doi:10.1136/gut.16.6.482
[4] Barnett, J.L., Hasler, W.L. and Camilleri, M. (1999)
American gastroenterological association medical posi-
tion statement on anorectal testing techniques. Gastroen-
terology, 116, 732-760.
doi:10.1016/S0016-5085(99)70194-0
[5] Bordeianou, L., Lee, K.Y., Rockwood, T., et al. (2008)
Anal resting pressures at manometry correlate with the
fecal incontinence severity index and with presence of
sphincter defects on ultrasound. Disease of Colon and
Rectum, 51, 1010-1014. doi:10.1007/s10350-008-9230-7
[6] Varma, M.G., Brown, J.S., Creasman, J.M., et al. (2006)
Fecal incontinence in females older than aged 40 years:
Who is at risk? Disease of Colon and Rectum, 49, 841-
851. doi:10.1007/s10350-006-0535-0
[7] Whitehead, W.E., Borrud, L., Goode, P.S., et al. (2009)
Fecal incontinence in US adults: Epidemiology and risk
factors. Gastroenterology, 137, 512-517.
doi:10.1053/j.gastro.2009.04.054
[8] Joh, H.K., Seong, M.K. and Oh, S.W. (2010) Fecal in-
continence in elderly Koreans. Journal of the American
Geriatrics Society, 58, 116-121.
doi:10.1111/j.1532-5415.2009.02613.x
[9] Lewis, S.J. and Heaton, K.W. (1997) Stool form scale as
a useful guide to intestinal transit time. Scandinavian
Journal of Gastroenterology, 32, 920-924.
doi:10.3109/00365529709011203
[10] Rao, S.S. and Patel, R.S. (1997) How useful are ma-
nometric tests of anorectal function in the management of

T. Abe et al. / Open Journal of Gastroenterology 3 (2013) 25-34 34
defecation disorders? American Journal of Gastroen-
terology, 92, 469-475.
[11] Abe, T., Kunimoto, M. and Hachiro, Y. (2008) Diagnosis
and management of fecal incontinence at a specialty out-
patient clinic. Journal of Japan Society of Coloproctology,
61, 247-253. doi:10.3862/jcoloproctology.61.247
[12] Laurberg, S. and Swash, M. (1989) Effects of aging on
the anorectal sphincters and their innervation. Disease of
Colon and Rectum, 32, 737-742.
doi:10.1007/BF02562120
[13] Enck, P., Kuhlbusch, R., Lübke, H., et al. (1989) Age and
sex and anorectal manometry in incontinence. Disease of
Colon and Rectum, 32, 1026-1030.
doi:10.1007/BF02553874
[14] Rao, S.S., Hatfield, R., Soffer, E., et al. (1999) Manomet-
ric tests in anorectal function in healthy adults. American
Journal of Gastroenterology, 94, 773-783.
doi:10.1111/j.1572-0241.1999.00950.x
[15] Gundling, F., Seidl, H., Scalercio, N., et al. (2009) Influ-
ence of gender and age on anorectal function: Normal
values from anorectal manometry in a large Caucasian
population. Digestion, 81, 207-213.
doi:10.1159/000258662
[16] Speakman, C.T.M., Hoyle, C.H.V., Kamm, M.A., et al.
(1995) Abnormal internal anal sphincter fibrosis and
elasticity in fecal incontinence. Disease of Colon and
Rectum, 38, 407-410. doi:10.1007/BF02054231
[17] Abe, T., Sato, Y., Kunimoto, M., et al. (2008) Effect of
aging and gender on internal anal sphincter thickness.
Anti-Aging Medicine, 5, 46-48. doi:10.3793/jaam.5.46
[18] MacLennan, A.H., Taylor, A.W., Wilson, D.H., et al.
(2000) The prevalence of pelvic floor disorders and their
relationship to gender, age, parity and mode of delivery.
British Journal of Obstetrics and Gynaecology, 107,
1460-1470. doi:10. 1111/j.1471-0528.2000.tb11669.x
[19] Nelson, R., Norton, N., Cautley, E., et al. (1995) Com-
munity-based prevalence of anal incontinence. Journal of
the American Medical Association, 274, 559-561.
doi:10.1001/jama.274.7.559
[20] Bharucha, A.E., Zinsmeister, A.R., Locke, G.R., et al.
(2006) Risk factors for fecal incontinence: A population-
based study in women. American Journal of Gastroen-
terology, 101, 1305-1312.
doi:10.1111/j.1572-0241.2006.00553.x
[21] Faltin, D.L., Boulvain, M., Irion, O., et al. (2000) Diag-
nosis of anal sphincter tears by postpartum endosonogra-
phy to predict fecal incontinence. Obstetrics and Gyne-
cology, 95, 643-647.
doi:10.1016/S0029-7844(99)00631-6
[22] Burnett, S.J., Spence-Jones, C., Speakman, C.T., et al.
(1991) Unsuspected sphincter damage following child-
birth revealed by anal endosonography. British Journal of
Radiology, 64, 225-227.
doi:10.1259/0007-1285-64-759-225
[23] Ward, A. and Tunuguntla, A.K. (1984) Anorectal sen-
sorimotor dysfunction in fecal incontinence and diabetes
mellitus. New England Journal of Medicine, 310, 1282-
1287. doi:10.1056/NEJM198405173102003
[24] Schiller, L.R., Santa Ana, C.A., Schmulen, C., et al.
(1982) Pathogenesis of fecal incontinence in diabetes
mellitus: Evidence for internal anal sphincter dysfunction.
New England Journal of Medicine, 307, 1666-1671.
doi:10.1056/NEJM198212303072702
[25] Rogers, J., Levy, D.M., Henry, M.M., et al. (1988) Pelvic
floor neuropathy: A comparative study of diabetes melli-
tus and idiopathic faecal incontinence. Gut, 29, 756-761.
doi:10.1136/gut.29.6.756
[26] Pintor, M.P., Zara, G.P., Falletto, E., et al. (1994) Pu-
dendal neuropathy in diabetic patients with faecal incon-
tinence. International Journal of Colorectal Disease, 9,
105-109. doi:10.1007/BF00699423
[27] Erckenbrecht, J.F., Winter, H.J., Cicmir, I., et al. (1988)
Faecal incontinence in diabetes mellitus: Is it correlated
to diabetic autonomic or peripheral neuropathy? Zeits-
chrift fur Gastroenterologie, 26, 731-736.
[28] Watanabe, M., Tsunoda, A., Kamiyama, G., et al. (2003)
Pathophysiology in diabetic patients with fecal inconti-
nence. Showa University Journal of Medical Sciences, 15,
21-26.
[29] Mills, K. and Chess-Williams, R. (2009) Pharmacology
of the internal anal sphincter and its relevance to faecal
incontinence. Autonomic and Autacoid Pharmacology, 29,
85-95. doi:10. 1111/j.1474-8673.2009.00437.x
[30] Yamato, S. and Rattan, S. (1990) Role of alpha adreno-
ceptors in opossum internal anal sphincter. Journal of
Clinical Investigation, 86, 424-429.
doi:10.1172/JCI114728
[31] Mendhekar, D.N., Srivastav, P.K., Sarin, S.K., et al.
(2003) A case report of olanzapine-induced fecal incon-
tinence. Journal of Clinical Psychiatry, 61, 601-602.
[32] Sagar, R., Varghese, S.T. and Balhara, Y.P. (2005) Olan-
zapine-induced double incontinesnce. Indian Journal of
Medical Sciences, 59, 163-164.
doi:10.4103/0019-5359.16123
[33] Siproudhis, L., Dinasquet, M., Sébille, V., et al. (2004)
Differential effects of two types of antidepressants, ami-
trptyline and fluoxetine, on anorectal motility and vis-
ceral perception. Alimentary Pharmacology and Thera-
peutics, 20, 689-695.
doi:10.1111/j.1365-2036.2004.02151.x
[34] Shamliyan, T., Wyman, J., Bliss, D.Z., et al. (2007) Pre-
vention of urinary and fecal incontinence in adults. Evi-
dence Reports/Technology Assessments, 161, 1-379.
[35] Harmston, C., Jones, O.M., Cunningham, C., et al. (2011)
The relationship between internal rectal prolapse and in-
ternal anal sphincter function. Colorectal Disease, 13,
791-795. doi:10. 1111/j.1463-1318.2010.02266.x
[36] Hiltunen, K.M., Matikainen, M., Auvinen, O., et al.
(1986) Clinical and manometric evaluation of anal sphinc-
ter function in patients with rectal prolapse. American
Journal of Surgery, 151, 489-492.
doi:10.1016/0002-9610(86)90110-8
[37] Furouk, R., Duthie, G.S., MacGregor, A.B., et al. (1994)
Recto-anal inhibition and incontinence in patients with
rectal prolapse. British Journal of Surgery, 81, 743-746.
doi:10.1002/bjs.1800810542