Abstract
Introduction and Hypothesis:
To investigate the relationship between race and anal incontinence (AI). Our hypotheses were (a) AI symptoms are similar between white and black women and (b) asymptomatic black and white women are equally likely to develop AI over one year of prospective observation.
Methods:
Parous women enrolled in a longitudinal cohort study were assessed for AI symptoms annually using Epidemiology of Prolapse and Incontinence Questionnaire. An AI score >0 indicated any bother from AI; a score >22.8 indicated clinically significant AI. We compared the odds of AI scores >0 at the visit level between white vs black women with logistic regression models using generalized estimating equations. We also estimated the odds of new AI symptoms at time T+1(one year later) among women free of AI symptoms at time T comparing white vs black women. In the latter analysis, we considered new AI symptoms to be represented by scores above 11.4. Covariates included in the adjusted models were: mode of delivery, obstetrical anal sphincter injuries, body mass index, age at the first delivery, and parity at enrollment.
Results.
Among 1256 participants, 189 (15.0%) were black. AI score= 0 was observed at 74.2% (= 5122/6902) person-visits. The adjusted odds ratio of AI score >0 was 1.83 (95% CI 1.24, 2.70) for white vs black women. Across 4364 visit pairs with AI score =0 at time T, 203 (4.7%) had AI score >11.4 at visit T+1 and white race significantly increased the odds of developing symptoms at time T+1 (adjusted OR= 2.26, 95% CI 1.28, 3.98).
Conclusions.
In an analysis that controlled for mode of delivery, obstetrical anal sphincter injuries, obesity, age at first delivery, and parity, white race was significantly associated with AI symptoms at any point in time as well as to the development of AI over one year of observation.
Keywords: Anal incontinence, AI score, Race
Brief Summary
In an analysis that controlled for mode of delivery, OASIS, obesity, age at delivery, and parity, white race was significantly associated with AI symptoms.
Introduction
Anal incontinence (AI), as defined by the International Urogynecological Association (IUGA) and the International Continence Society (ICS), is a “complaint of involuntary loss of feces or flatus”[1]. Definitions used for epidemiologic studies vary and most of the studies only include fecal incontinence (FI), which is defined as a “complaint of involuntary loss of feces” [1]. Depending on the definition used, the prevalence of fecal incontinence reported varies from 2.2% to 24% in women in the United States [2–5]. Several studies have investigated risk factors for AI and established risk factors include age, obesity, vaginal delivery, and obstetric anal sphincter injuries (OASIS) [4, 6–10].
The relationship between race and AI is unclear. In the literature in which race is included as a covariate, the association of race with AI is inconsistent. Some studies reported that AI is not significantly associated with race or ethnicity [4, 11–13], while others reported fecal incontinence was less common in black women compared with white women [14, 15]. Study populations were different, and they usually lacked adjustment for confounders. In addition, most studies investigating AI are cross-sectional. Very little is known about the incidence of AI, or about the change in this disease over time. Therefore, in this study, we used data from a longitudinal cohort study of parous women to assess the relationship between race and AI. We investigated two hypotheses: (a) black and white women are equally likely to have AI symptoms; (b) asymptomatic black and white women are equally likely to develop AI over one year.
Methods and Materials
Study setting and population
This study is an analysis of the data collected by The Mothers’ Outcomes After Delivery (MOAD) study. The MOAD study was a prospective longitudinal cohort study of pelvic floor outcomes in 1529 women enrolled between October 2008 and November 2013. This study was designed and conducted by investigators from the Johns Hopkins Medical Institutions and the Greater Baltimore Medical Center in Baltimore, Maryland. It received Institutional Review Board Approval at both institutions and all participants provided written informed consent.
A primary goal of the MOAD study was to describe and compare the incidence of pelvic floor disorders between parous women with different modes of delivery [16]. The study recruited participants who delivered their first baby at Great Baltimore Medical Center. Detailed recruitment methods, selection criteria and exclusion criteria have been described previously [17]. Briefly, recruitment was based on delivery type (cesarean prior to labor, cesarean during labor, vaginal birth). Diagnoses from obstetric hospital discharge records were reviewed by trained personnel to identify eligible participants and verify the delivery type. Participants were matched based on delivery mode, participant age (in five-year strata) as well as number of years from the first delivery (in ¼ -year strata). After matching, potential participants were screened for eligibility via telephone interview as well. By design, this study had high proportion of women who delivered by cesarean (around 50%).
After enrollment, participants were followed prospectively and were assessed annually for pelvic floor symptoms and prolapse. The follow-up stopped on April 2017. Therefore, at the time of analysis, a participant might have participated in 9 annual assessments at the most.
For this present study, since the aim was to investigate changes over time, we excluded women who, in the parent study (MOAD), only had one visit (n=207).
Exposures
The primary exposure, race, was derived from the enrollment data. It was self-reported, and participants could choose from these categories: American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, or White. Since our hypothesis pertained only to white vs black race, we excluded 64 women who reported any other race and two with missing race yielding 1256 women in our study population.
Confounders considered included: mode of delivery, body mass index (BMI), age at the first delivery and parity at enrollment. Delivery mode was classified as cesarean versus vaginal birth. Furthermore, since obstetrical anal sphincter injuries (OASIS) is a risk factor for developing AI, we collected data on anal sphincter injury from medical records. Participants who ever had a third- or fourth-degree tear were considered to have OASIS. Thus, considering all deliveries for each participant, participants were grouped into three categories: cesarean birth only, at least one vaginal birth but no history of OASIS, and at least one vaginal birth complicated by OASIS.
At each annual visit, weight and height were measured, and BMI was calculated (kg/m2). In this analysis, BMI was time-varying and categorical with three categories: <25 kg/m2, 25-30 kg/m2, and ≥ 30 kg/m2. BMI≥ 30 kg/m2 was defined as obesity and 25-30 kg/m2 as overweight. Parity at enrollment and age at the first delivery were obtained by self-report at study enrollment.
Outcomes
At each visit, each participant completed the Epidemiology of Prolapse and Incontinence Questionnaire (EPIQ). EPIQ was developed to identify pelvic floor disorders and has been rigorously validated in a community population [18]. At each annual visit, each participant was asked, “Do you lose gas from your rectum that is beyond your control?” If she answered yes, she then was asked: “How much are you bothered by losing gas from your rectum?” The answer would be a score between 0 and 100 inclusive, where 100 indicated most bothersome. This score became the score for gas. Similarly, scores for liquid and solid were also obtained. The AI score was computed from the average of these three scores. A cutoff point of >22.8 had been identified to maximize the positive predictive value for identifying women with clinically significant symptoms of AI (sensitivity 87, specificity 70) [18]. In addition, test-retest reliability for measures of the degree of bother related to AI were good (correlations of 0.70 or above, except for the question ‘How bothered by uncontrolled solid stool loss’ whose intraclass correlation coefficient= 0.51) [18]. Since AI score was quantitative and had a validated threshold for clinical AI, we used it to measure the presence of and severity of AI. More specifically, a score of 0 was equivalent to having no AI symptoms or having symptoms with absolutely no bother. A score greater than 22.8 was equivalent to burdensome or clinically significant AI. A score between 0 and 22.8 corresponded to AI symptoms with a level of bother that did not meet the threshold for clinical significance. In this analysis, we considered a meaningful increase as AI score increase more than 11.4 (half of 22.8) over one year of follow-up (i.e. AI score at time T+1 being greater than 11.4). Since EPIQ was self-administered each year, AI scores could change per year.
Statistical analysis
Our first hypothesis was that black and white women were equally likely to have AI symptoms (e.g., AI score>0). For this hypothesis, we used a logistic regression model to estimate the probability of AI score being >0 by race and generalized estimating equation (GEE) methods to incorporate the statistical dependences of repeated assessment of AI within women.
For our second hypotheses, the analysis was at the level of a visit pair (two sequential visits from any participant, at time T and time T+1). In total, 5872 visit pairs which were 1.08+0.29 years apart were derived from 6902 visits. Of them, 4364 had AI score at time T equal to zero. We used a logistic regression model with GEE to estimate the odds ratio (OR) of AI score at visit T+1 being greater than 11.4 for white women vs. black women among those free of symptoms at time T (AI scores at visit T= 0). The formula of the regression model was
log (odds of AI score at time T+1 > 11.4│ AI score at time T= 0) = β0 +β1 if white + α2 if vaginal without OASIS + α3 if vaginal with OASIS + β4 if (25kg/m2 ≤ BMI at time T<30kg/m2) + α5 if (BMI≥ 30kg/m2) + α6(Age at first delivery-25) + α7 if (parity at enrollment= 2) + β8 if (parity at enrollment≥ 3).
Hence, exp(α0) represents the odds of AI for a black women, who only had cesarean delivery, whose BMI was < 25kg/m2, was 25 years old at her first delivery, and have had only one delivery at enrollment.
Results
Overall, 1529 women were enrolled in the parent study. We excluded 207 participants who only came for one visit, 64 who reported being “other race” and two who did not report race, leaving 1256 participants with 7245 visits, which provided 5989 visit pairs. However, BMI data were missing for some visits, leaving 6902 visits and 5872 visit pairs for the analysis.
Table 1 shows the characteristics of the study population at enrollment. Compared to black women, white women were older when they had their first delivery (32.3 vs 29.7 years old, P< 0.001) and entered into this study later after their first delivery (7.1 vs 6.7 years, P= 0.001). Black women were more likely to have a cesarean delivery (59.3% vs 48.5%, P= 0.003) and were more likely to be obese (47.6% vs 21.3%, P< 0.001). White women were more likely to have high parity: about 17.8% of white women had three deliveries or more at enrollment compared to only 7.4% among black women. White women attended more visits than black women but there was no statistically significant difference. At enrollment, absence of AI symptoms was higher among black women (85.7% vs 73.4%, P= 0.002). Furthermore, 12.7% of white women had AI scores greater than 22.8 at enrollment, about three times the proportion among black women. (Table 1)
Table 1.
Characteristics of the study population at the first study visit (i.e. enrollment)
| Characteristics | Black (n=189) | White (n=1067) | p-value | |
|---|---|---|---|---|
| Age at first delivery, mean (SD) | 29.7 (5.6) | 32.3 (4.7) | <0.001 | |
| Mode of delivery | ||||
| Cesarean | 112 (59.3) | 517 (48.5) | 0.003 | |
| Vaginal without OASIS | 70 (37.0) | 444 (41.6) | ||
| Vaginal with OASIS | 7 (3.7) | 106 (9.9) | ||
| BMI at enrollment, kg/m2 | ||||
| <25 | 41 (21.7) | 541 (50.7) | <0.001 | |
| 25-30 | 58 (30.7) | 299 (28.0) | ||
| >=30 | 90 (47.6) | 227 (21.3) | ||
| Parity at enrollment | ||||
| 1 | 83 (43.9) | 265 (24.8) | <0.001 | |
| 2 | 92 (48.7) | 612 (57.4) | ||
| >=3 | 14 (7.4) | 190 (17.8) | ||
| Years since 1st delivery at enrollment, mean (SD) | 6.7 (1.7) | 7.1 (1.7) | 0.001 | |
| No. of follow-up visits | ||||
| 2-3 | 39 (20.6) | 206 (19.3) | 0.13 | |
| 4-5 | 50 (26.5) | 220 (20.6) | ||
| 6-7 | 60 (31.7) | 339 (31.8) | ||
| 8-9 | 40 (21.2) | 302 (28.3) | ||
| AI score at enrollment | ||||
| =0 | 162(85.7) | 783(73.4) | 0.002 | |
| 0-11.4 | 11(5.8) | 94(8.8) | ||
| 11.4-22.8 | 8(4.2) | 55(5.1) | ||
| >22.8 | 8(4.2) | 135(12.7) | ||
Data are presented as No. (percentage) unless otherwise indicated.
At the visit level, the distribution of AI scores for each visit was highly right-skewed. Of 6902 visits, 5122 (74.2%) had an AI score equal to 0, corresponding to either no AI symptoms or AI symptoms without any bother. Among those scores greater than zero, 630 (9.1%) were greater than 0 but lower than 11.4, 386 (5.6%) were between 11.4 and 22.8, and 764 (11.1%) were greater than 22.8.
Our first hypothesis was that AI symptoms are similar between white and black women. This analysis was performed at the visit level (Table 2). Among black women, an AI score of >0 was reported in 154 visits of 997 visits (15.5%). Among white women, an AI score of >0 was reported in 1626 of 5905 visits (27.5%). Thus, the odds of AI score being greater than 0 among white women was 2.08 times the odds among black women and was statistically significant (95% CI 1.44, 3.01). After adjusting for mode of delivery, BMI, age at the first delivery and parity at enrollment, the odds ratio of AI score>0 was 1.83 (95% CI 1.24, 2.70) for white vs black women. Other covariates associated with AI symptoms in the multivariable model included vaginal delivery versus cesarean (OR 1.34, 95% CI 1.06, 1.71), OASIS (OR 2.16, 95% CI 1.49, 3.14), obesity (OR 1.45, 95% CI 1.12, 1.88), and age at the first delivery (OR 1.06 per one-year increase, 95% CI 1.03, 1.09). Since 22.8 was the cut-off point for AI score to diagnose clinically significant anal incontinence in this questionnaire, we also performed a separate analysis to estimate the odds ratio for AI score >22.8. After adjusting for mode of delivery, BMI, age at delivery and deliveries at enrollment, white race increased the risk by almost 2.5 (OR 2.48, 95% CI 1.44, 4.27).
Table 2.
Univariable and multivariable logistic regression models using GEE for AI score>0 among 6902 visits in 1256 women.
| Univariable OR (95% CI) | Multivariable* OR (95% CI) | ||
|---|---|---|---|
| Race | |||
| Black (n= 997) | reference | reference | |
| White (n= 5905) | 2.08 (1.44-3.01) | 1.83 (1.24-2.70) | |
| Mode of delivery | |||
| Cesarean (n= 3536) | reference | reference | |
| Vaginal without OASIS (n=2743) | 1.30 (1.03-1.64) | 1.34 (1.06-1.71) | |
| Vaginal with OASIS (n= 623) | 2.12 (1.47-3.06) | 2.16 (1.49-3.14) | |
| BMI (kg/m2) | |||
| <25 (n= 2976) | reference | reference | |
| 25-30 (n= 2008) | 0.97 (0.78-1.21) | 1.10 (0.88-1.38) | |
| >=30 (n= 1918) | 1.05 (0.82-1.34) | 1.45 (1.12-1.88) | |
| Age at the first delivery-25 (years) | 1.05 (1.03-1.08) | 1.06 (1.03-1.09) | |
| Parity at enrollment | |||
| 1 (n= 1806) | reference | reference | |
| 2 (n= 3872) | 0.97 (0.75-1.25) | 1.12 (0.85-1.48) | |
| >=3 (n= 1224) | 1.14 (0.81-1.58) | 1.35 (0.93-1.94) | |
Abbreviations: AI, anal incontinence; OR, odds ratio; CI, confidence interval; OASIS, obstetric anal sphincter injuries; BMI, body mass index
adjusting for mode of delivery, BMI, age at the first delivery and parity at enrollment.
Visit pairs were analyzed for our second hypotheses, which was that white and black women who are free of AI symptoms at time T are equally likely to develop new AI symptoms at time T+1. Of 5872 visit pairs, 3965 (67.5%) were 0-0 pairs (AI score at visit T and T+1 were both 0). Only 399 (6.8%) visit pairs developed AI symptoms at time T+1 when they had no symptoms at time T. Among these 399 visit pairs, 203 (50.9%) had an AI score greater than 11.4 at time T+1. For all visit pairs in which the score was zero at time T (4364 visit pairs), the proportion of visit pairs with a score greater than 11.4 at time T+1 was less for blacks than for whites (18 (2.5%) versus 185 (5.1%), P= 0.004) (Figure 1). As shown in Table 3, white race significantly increased the odds to develop the symptoms at time T+1 (OR 2.04, 95% CI 1.19, 3.52) which after adjusting for mode of delivery, BMI at visit T, age at the first delivery and parity at enrollment resulted in an OR= 2.26 (95% CI 1.28, 3.98). Other covariates associated with the development of AI symptoms included vaginal delivery without OASIS (OR 1.68, 95% CI 1.20, 2.35) and vaginal delivery with OASIS (OR 1.88, 95% CI 1.05, 3.35). In addition, among 4364 visit pairs in which the score was zero at time T, there were 9 AI score at time T+1 greater than 22.8 among black women (1.27%) whereas there were 100 among white women (2.74%) which yielded an odds ratio of 2.18 which was consistent with the one using 11.4 as the cut off value.
Figure 1.
AI score at visit T+1 by race for visit pairs in which AI score at visit T=0 (n= 4364 visit pairs) (*Proportion of visit pairs in which AI score at visit T+1 was zero)
Table 3.
Univariable and multivariable logistic regression models using GEE for new onset AI, defined as AI score at visit T+1 greater than 11.4 among visit pairs in which AI score at visit T was equal to 0 (n= 4364 visit pairs).
| Univariable OR (95% CI) | Multivariable* OR (95% CI) | ||
|---|---|---|---|
| Race | |||
| Black (n= 708) | reference | reference | |
| White (n= 3656) | 2.04 (1.19-3.52) | 2.26 (1.28-3.98) | |
| Mode of delivery | |||
| Cesarean (n= 2361) | reference | reference | |
| Vaginal without OASIS (n= 1677) | 1.64 (1.18-2.30) | 1.68 (1.20-2.35) | |
| Vaginal with OASIS (n= 326) | 1.84 (1.04-3.28) | 1.88 (1.05-3.35) | |
| BMI at visit T (kg/m2) | |||
| <25 (n= 1924) | reference | reference | |
| 25-30 (n= 1264) | 0.92 (0.63-1.35) | 1.02 (0.69-1.50) | |
| >=30 (n= 1176) | 1.11 (0.76-1.61) | 1.44 (0.98-2.12) | |
| Age at the first delivery-25 (years) | 1.00 (0.97-1.03) | 1.00 (0.96-1.03) | |
| Parity at enrollment | |||
| 1 (n= 1153) | reference | reference | |
| 2 (n= 2474) | 1.05 (0.72-1.52) | 0.97 (0.66-1.44) | |
| >=3 (n= 737) | 1.02 (0.64-1.65) | 0.86 (0.52-1.43) | |
Abbreviations: AI, anal incontinence; OR, odds ratio; CI, confidence interval; OASIS, obstetric anal sphincter injuries; BMI, body mass index
adjusting for mode of delivery, BMI at visit T, age at the first delivery and parity at enrollment.
Discussion
An important finding from our study is that parous black women are less likely to report any AI symptoms than parous white women. In this population, the odds of reporting any AI symptoms were about two times higher for white women. The other important result we reported here is that among those free of AI symptoms at time T, black women are also less likely to develop de novo AI symptoms at time T+1 than white women.
Although there is paucity of data on the racial disparity in AI and the potential reasons, there are some studies reporting racial disparity in urinary incontinence and pelvic organ prolapse [19, 20]. Analyses of the Reproductive Risks for Incontinence Study at Kaiser showed that UI and symptomatic POP were less common among African-Americans [21, 22]. Those results, together with our findings, suggest that black women may at lower risk to experience pelvic floor disorders. It is not clear whether this differences might be due to possible racial variations in anatomy and connective tissue quality. Van Dongen proposed some factors in his 1982 report to explain this disparity. In addition to the observed deeper and smaller pelvis in black women, he also commented on the importance of the stronger and tougher supportive fascia and ligaments in the black than in the white [23]. Downing et al compared the thickness of levator and obturator muscles between asymptomatic black and white nulliparas using three-dimensional magnetic resonance imaging (MRI) color mapping, observing significantly thicker levator muscles in black women [24]. In addition, a study measuring the bony pelvis demonstrated a smaller posterior and total pelvic area in the African-American as compared to white women [25]. However, another study of 234 participants having MRIs at 6–12 months postpartum reported that there were no significant differences between African-American women and white women in interspinous diameter, angle of the subpubic arch, anteroposterior conjugate, levator thickness, or levator hiatus [26]. Thus, it is impossible to conclude whether our findings might be attributed to inherent differences in pelvic floor anatomy or connective tissue quality by race; more studies are needed to investigate the racial difference in anatomy and histology of the pelvic muscles and supportive structures.
Another explanation for the finding might be the subjective value of “bothersome”. It is possible that black women value symptom bothers differently than white women. The study population for the validation of the Epidemiology of Prolapse and Incontinence Questionnaire included a relatively small proportion of African-Americans. It is possible that race (or culture) could modify the relationship between the questionnaire score and the underlying construct of AI severity. However, no studies have investigated this, and we cannot be sure that black women responded differently from white women. Other explanations for our findings might be unmeasured confounders, including medication use, diet, or other bowel diseases.
Several other findings from our study are also worth highlighting. First, vaginal delivery is highly related to the increased risk of anal incontinence. This finding is similar to many other studies [12, 27]. In addition, vaginal deliveries with OASIS and odds of AI score above zero showed a dose-response relationship in Table 2. This corresponds to another published analysis of MOAD data, which reported that operative vaginal delivery was associated with significantly higher hazard of AI (adjusted hazard ratio, 1.75 [95% CI, 1.14-2.68]) [28]. Additionally, some cross-sectional studies also reported the same results [29, 30].
Our research has several limitations. First, our analysis only focused on the comparison between black and white, two major races in the United States, so our findings are limited to these two race groups. Second, we used 11.4 as the cutoff point to define a meaningful increase in AI score and the strength of the magnitude of the association was confirmed when we used the cutoff of 22.8 but the number of events among the black women was small (n= 9). Finally, this study was done at only one site, therefore, may not be generalizable to other populations.
Our study had several strengths. First, the parent study was in longitudinal design which is rare for pelvic floor disorders. It allowed us to investigate long-term AI outcome whereas most other studies were cross-sectional or only focused on short-term outcomes. With the longitudinal data, we also were able to incorporate more information into analysis and investigate the change in AI score across one year. Second, since our main exposure in this analysis was collected at baseline and only pertained to white and black race, misclassification was very unlikely. Third, we were able to control for several confounders including mode of delivery, BMI, age and parity, and models we used were able to account for time-varying measures.
In conclusion, white race is associated with increased risk of AI symptoms. Compared with black women, white women are more likely to have AI symptoms or report any AI symptoms if they are free of symptoms one year before.
Acknowledgments
Jennifer Roem deserves acknowledgment for data set preparation.
Funding
This study was funded by grants R01HD082070 and R01HD056275 from Eunice Kennedy Shriver National Institute of Child Health and Human Development.
Footnotes
Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of a an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.
Financial Disclaimers/Conflict of Interest
Financial Disclaimers/Conflict of Interest: NONE
References
- 1.Sultan AH, Monga A, Lee J, Emmanuel A, Norton C, Santoro G, Hull T, Berghmans B, Brody S, Haylen BT (2017) An International Urogynecological Association (IUGA)/International Continence Society (ICS) joint report on the terminology for female anorectal dysfunction. Neurourol Urodyn 36:10–34 [DOI] [PubMed] [Google Scholar]
- 2.Sung VW, Hampton BS (2009) Epidemiology of pelvic floor dysfunction. Obstet Gynecol Clin North Am 36:421–443 [DOI] [PubMed] [Google Scholar]
- 3.Shamliyan T, Wyman J, Bliss DZ, Kane RL, Wilt TJ (2007) Prevention of urinary and fecal incontinence in adults. Evid Rep Technol Assess (Full Rep) 1–379 [PMC free article] [PubMed] [Google Scholar]
- 4.Whitehead WE, Borrud L, Goode PS, Meikle S, Mueller ER, Tuteja A, Weidner A, Weinstein M, Ye W, Pelvic Floor Disorders Network (2009) Fecal incontinence in US adults: epidemiology and risk factors. Gastroenterology 137:512–517, 517.e1–2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Varma MG, Brown JS, Creasman JM, Thom DH, Van Den Eeden SK, Beattie MS, Subak LL, Reproductive Risks for Incontinence Study at Kaiser (RRISK) Research Group (2006) Fecal incontinence in females older than aged 40 years: who is at risk? Dis Colon Rectum 49:841–851 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ditah I, Devaki P, Luma HN, Ditah C, Njei B, Jaiyeoba C, Salami A, Ditah C, Ewelukwa O, Szarka L (2014) Prevalence, Trends, and Risk Factors for Fecal Incontinence in United States Adults, 2005–2010. Clinical Gastroenterology and Hepatology 12:636–643.e2 [DOI] [PubMed] [Google Scholar]
- 7.Gyhagen M, Bullarbo M, Nielsen TF, Milsom I (2014) Faecal incontinence 20 years after one birth: a comparison between vaginal delivery and caesarean section. Int Urogynecol J 25:1411–1418 [DOI] [PubMed] [Google Scholar]
- 8.Erekson EA, Sung VW, Myers DL (2008) Effect of body mass index on the risk of anal incontinence and defecatory dysfunction in women. Am J Obstet Gynecol 198:596.e1–4 [DOI] [PubMed] [Google Scholar]
- 9.Altman D, Ekström A, Forsgren C, Nordenstam J, Zetterström J (2007) Symptoms of anal and urinary incontinence following cesarean section or spontaneous vaginal delivery. Am J Obstet Gynecol 197:512.e1–7 [DOI] [PubMed] [Google Scholar]
- 10.Soerensen MM, Buntzen S, Bek KM, Laurberg S (2013) Complete obstetric anal sphincter tear and risk of long-term fecal incontinence: a cohort study. Dis Colon Rectum 56:992–1001 [DOI] [PubMed] [Google Scholar]
- 11.Goode PS, Burgio KL, Halli AD, Jones RW, Richter HE, Redden DT, Baker PS, Allman RM (2005) Prevalence and correlates of fecal incontinence in community-dwelling older adults. J Am Geriatr Soc 53:629–635 [DOI] [PubMed] [Google Scholar]
- 12.Nygaard I (2008) Prevalence of Symptomatic Pelvic Floor Disorders in US Women. JAMA 300:1311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Melville JL, Fan M-Y, Newton K, Fenner D (2005) Fecal incontinence in US women: a population-based study. Am J Obstet Gynecol 193:2071–2076 [DOI] [PubMed] [Google Scholar]
- 14.Townsend MK, Matthews CA, Whitehead WE, Grodstein F (2013) Risk factors for fecal incontinence in older women. Am J Gastroenterol 108:113–119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Burgio KL, Borello-France D, Richter HE, et al. (2007) Risk factors for fecal and urinary incontinence after childbirth: the childbirth and pelvic symptoms study. Am J Gastroenterol 102:1998–2004 [DOI] [PubMed] [Google Scholar]
- 16.Handa VL, Pierce CB, Muñoz A, Blomquist JL (2015) Longitudinal changes in overactive bladder and stress incontinence among parous women. Neurourol Urodyn 34:356–361 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Handa VL, Blomquist JL, Knoepp LR, Hoskey KA, McDermott KC, Muñoz A (2011) Pelvic floor disorders 5-10 years after vaginal or cesarean childbirth. Obstet Gynecol 118:777–784 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lukacz ES, Lawrence JM, Buckwalter JG, Burchette RJ, Nager CW, Luber KM (2005) Epidemiology of prolapse and incontinence questionnaire: validation of a new epidemiologic survey. Int Urogynecol J Pelvic Floor Dysfunct 16:272–284 [DOI] [PubMed] [Google Scholar]
- 19.Danforth KN, Townsend MK, Lifford K, Curhan GC, Resnick NM, Grodstein F (2006) Risk factors for urinary incontinence among middle-aged women. Am J Obstet Gynecol 194:339–345 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Knobel J (1975) Stress incontinence in the Black female. S Afr Med J 49:430–432 [PubMed] [Google Scholar]
- 21.Rortveit G, Brown JS, Thom DH, Van Den Eeden SK, Creasman JM, Subak LL (2007) Symptomatic pelvic organ prolapse: prevalence and risk factors in a population-based, racially diverse cohort. Obstet Gynecol 109:1396–1403 [DOI] [PubMed] [Google Scholar]
- 22.Thom DH, van den Eeden SK, Ragins AI, Wassel-Fyr C, Vittinghof E, Subak LL, Brown JS (2006) Differences in Prevalence of Urinary Incontinence by Race/Ethnicity. The Journal of Urology 175:259–264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.van Dongen L (1981) The anatomy of genital prolapse. S Afr Med J 60:357–359 [PubMed] [Google Scholar]
- 24.Downing KT, Hoyte LP, Warfield SK, Weidner AC (2007) Racial differences in pelvic floor muscle thickness in asymptomatic nulliparas as seen on magnetic resonance imaging-based three-dimensional color thickness mapping. Am J Obstet Gynecol 197:625.e1–4 [DOI] [PubMed] [Google Scholar]
- 25.Baragi RV, DeLancey JOL, Caspari R, Howard DH, Ashton-Miller JA (2002) Differences in pelvic floor area between African American and European American women. American Journal of Obstetrics and Gynecology 187:111–115 [DOI] [PubMed] [Google Scholar]
- 26.Handa VL, Lockhart ME, Fielding JR, Bradley CS, Brubaker L, Cundiff GW, Ye W, Richter HE, Pelvic Floor Disorders Network (2008) Racial differences in pelvic anatomy by magnetic resonance imaging. Obstet Gynecol 111:914–920 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Pretlove SJ, Thompson PJ, Toozs-Hobson PM, Radley S, Khan KS (2008) Does the mode of delivery predispose women to anal incontinence in the first year postpartum? A comparative systematic review. BJOG 115:421–434 [DOI] [PubMed] [Google Scholar]
- 28.Blomquist JL, Muñoz A, Carroll M, Handa VL (2018) Association of Delivery Mode With Pelvic Floor Disorders After Childbirth. JAMA 320:2438–2447 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Turel F, Langer S, Shek KL, Dietz HP (2018) Medium- to Long-term Follow-up of Obstetric Anal Sphincter Injury. Dis Colon Rectum, doi: 10.1097/DCR.0000000000001297 [DOI] [PubMed] [Google Scholar]
- 30.Schei B, Johannessen HH, Rydning A, Sultan A, Mørkved S (2019) Anal incontinence after vaginal delivery or cesarean section. Acta Obstet Gynecol Scand 98:51–60 [DOI] [PubMed] [Google Scholar]