If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Reprint requests Address requests for reprints to: Ivo Ditah, MD, MPhil, Division of Gastroenterology and Hepatology, Mayo Clinic and Foundation, 200 First Street SW, Rochester, Minnesota 55905. fax: (507) 284-0538.
Department of Medicine, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MinnesotaClinical Enteric Neurosciences Translational and Epidemiology Research Program, Mayo Clinic, Rochester, Minnesota
We investigated the prevalence of and trends and risk factors for fecal incontinence (FI) in the United States among non-institutionalized adults from 2005 to 2010.
We analyzed data from 14,759 participants in the U.S. National Health and Nutrition Examination Survey (49% women, 20 years or older) from 2005 to 2010 (the FI Severity Index was added in 2005–2006). FI was defined as accidental leakage of solid or liquid stool or mucus at least once in preceding month. Sampling weights were used to obtain estimates for the national population. Logistic regression was used to identify risk factors for FI.
The prevalence of FI among non-institutionalized U.S. adults was 8.39% (95% confidence interval, 7.76–9.05). It was stable throughout the study period: 8.26% in 2005–2006, 8.48% in 2007–2008, and 8.41% in 2009–2010. FI resulted in release of liquid stool in most cases (6.16%). Prevalence increased with age from 2.91% among 20- to 29-year-old participants to 16.16% (14.15%–18.39%) among participants 70 years and older. Independent risk factors for FI included older age, diabetes mellitus, urinary incontinence, frequent and loose stools, and multiple chronic illnesses. FI was more common among women only when they had urinary incontinence.
FI is a common problem among non-institutionalized U.S. adults. Its prevalence remained stable from 2005–2010. Diabetes mellitus and chronic diarrhea are modifiable risk factors. Future studies on risk factors for FI should assess for presence of urinary incontinence.
In 2005, the Pelvic Floor Disorders Network submitted a validated FI severity scale for inclusion in the National Health and Nutrition Evaluation Survey (NHANES). The first national prevalence of FI was reported in 2009.
These estimates were limited by small number of participants.
From a public health perspective, the burden of FI on individuals, families, and society cannot be underestimated. FI is projected to become a major challenge as the proportion of the elderly increases in the general population. Age remains the most consistent risk factor for FI.
included only 4308 individuals. Our study involved 14,759 participants, making it the largest ever population-based study on FI in the United States. Our study involved 3 NHANES cycles, allowing for the assessment of FI trends.
To date, there are no data on the trends of FI in the United States. We hypothesized that because of the changing age distribution of the U.S. population, the prevalence of FI is almost certainly on the increase. The NHANES has collected and reported data on FI from 2005 through 2010. This represents the largest data set on FI and allows for calculation of robust estimates. The aims of this study were to (1) estimate the prevalence of FI, (2) analyze trends in FI, and (3) identify independent risk factors for FI among U.S. adults.
The NHANES program consists of annual cross-sectional, national health surveys conducted by the National Center for Health Statistics, Centers for Disease Control and Prevention. Demographic, socioeconomic, and health interview data are collected in the home, followed by physical examinations, interviews, and laboratory assessments in mobile examination centers.
The detailed procedural methods pertaining to data on FI have been previously published.
which asks about the frequency of accidental bowel leakage during the past month separately for gas, mucus, liquid, and solid stool. Frequency is assessed as 2 or more times a day, once a day, 2 or more times a week, once a week, 1–3 times a month, or never. For this study, FI was defined as any involuntary loss of mucus, liquid, or solid stool in the last 30 days. The bowel health questionnaire also asked survey participants how often they usually have bowel movements. For data analysis, their responses were merged into 3 ranges: <3/wk, 3–21/wk, and >21/wk.
For the NHANES survey, participants were asked, “What is your usual or most common stool type?” For this study, we pooled stool type ratings 1 and 2 (hard and lumpy), ratings 3–5 (normal consistencies), and ratings 6 and 7 (mushy and watery).
Table 1 summarizes definitions of the various factors included in the study.
Table 1Definitions and/or Categorization of Operational Terms
Only individuals older than 20 years were eligible. Age was categorized as 20–29 years, 30–39 years, 40–54 years, 55–69 years, and 70 years and older.
On the basis of self-reported information, participants were classified into Mexican American and other Hispanic, non-Hispanic white, non-Hispanic black, and other race/ethnicity groups.
It was classified as widowed, separated, or divorced; married or cohabitating with a partner; and never married.
Education was defined as having less than a high school education, a high school diploma (including general educational development [GED]), or additional education beyond high school.
The body mass index (BMI) was calculated as measured weight (kg)/height (m2); overweight was defined by a BMI of 25.0–29.9 kg/m2, and obesity was defined by a BMI >30 kg/m2.
Vigorous physical activity was defined as activity done for at least 10 minutes in the past 30 days that caused heavy sweating or large increases in breathing or heart rate; moderate physical activity was defined as activity done for at least 10 minutes that caused only light sweating or a slight to moderate increase in breathing or heart rate.
Pregnancy status was assessed by participant self-report and/or a urine pregnancy test.
Diabetes mellitus was defined as a positive response to any of 3 questions: (1) “Have you ever been told by a doctor or other health professional that you have diabetes or sugar diabetes?”, (2) “Are you now taking insulin?”, and (3) “Are you now taking diabetic pills to lower your blood sugar?”
The presence of “Chronic illnesses” was ascertained by 13 separate questions on the form: “Have you ever been told by a doctor that you have arthritis, congestive heart failure, coronary heart disease, angina or angina pectoris, heart attack, stroke, emphysema, chronic bronchitis, chronic liver condition, cancer or malignancy, asthma, anemia, visual problems, or osteoarthritis?” The cumulative number of positive responses to these 13 questions was divided into 3 categories: 0 (none), 1–2, and 3 or more.
General health status
Self-described general health status was defined by the question “Would you say that in general your health is excellent, very good, good, fair, or poor?” Responses to this question were aggregated into 2 ranges: excellent, very good, or good health status versus fair or poor health.
Urinary incontinence was assessed by using the Incontinence Severity Index,
which consists of 2 questions: “How often do you have urinary leakage? Would you say never, less than once a month, a few times a month, a few times a week, or every day and/or night (responses coded 1–5)?” and “How much urine do you lose each time? Would you say drops, small splashes, or more (responses coded 1–3)?” This questionnaire is scored by multiplying responses to the 2 questions, and scores of 3 or greater are classified as at least moderately severe urinary incontinence.
All estimates and measures of association were derived by using the sampling weights provided by the National Center for Health Statistics. These weights consider unequal probabilities of selection resulting from sample design, nonresponse, and planned oversampling of the elderly, non-Hispanic black, and Mexican-American populations.
The analysis was limited to adults 20 years and older. Individuals with missing questionnaire items or who refused to answer the FI questions were excluded from the analysis. Stata program (Stata Corporation, College Station, TX) was used to calculate prevalence estimates and their 95% confidence intervals (CIs). The examination sample weights adjust for unequal probabilities of selection and nonresponse and are post-stratified to population control totals for each sampling subdomain. Estimates with relative standard errors greater than 30% are identified as statistically unreliable. Prevalence was stratified by age, gender, and stool type. Time trends are presented by age and gender. The results are shown on bar charts and tables. First, age-adjusted bivariate analyses were performed. Only factors that reached a significance limit of 0.10 in this analysis were included in the multivariate analysis. The model was built in a forward manner by adding one variable at a time. Interaction between UI, gender, and age to explore for effect modification was also performed. The Pearson χ2 test with a significance limit of 0.05 was used to identify risk factors for FI.
Overall, 23,198 participants 20 years and older were invited to participate in the NHANES household survey during the 5-year period. Of these, 16,539 individuals (71.3%) responded to the invitation and completed the survey. Unfortunately, 1776 of the respondents (10.7%) did not complete the health examination components that included the bowel health questionnaire. Four participants (0.03%) specifically refused to complete the FI section of the questionnaire. For this study, only the 14,759 participants who completed the FI questionnaire have been included in the analysis. These participants were distributed as follows: 4308 (29.2%) in 2005–2006, 5 174 (35.1%) in 2007–2008, and 5277 (35.8%) in 2009–2010. Supplementary Figure 1 shows flowchart of survey invitees and response during each study cycle.
Demographic Characteristics of Survey Participants
The study population was made up of 7511 women (51.3%) and 7248 men (48.7%). The mean (and median) age of the study participants was 49 years (standard deviation, 18). All age groups were well-represented and mirrored the overall U.S. population. Hispanic subjects represented 13.4%, and non-Hispanic black subjects were 10.18% of study population. Supplementary Table 1 shows the demographic characteristics of all participants.
Prevalence of Fecal Incontinence
The overall prevalence of FI among civilian non-institutionalized U.S. adults 20 years and older was 8.39% (95% CI, 7.76–9.05). This represents a total of about 19 million non-institutionalized individuals in the United States who report at least 1 episode of FI during the last 30 days as of 2010. The prevalence of FI was higher in women (9.4%; 95% CI, 8.8–10.27) than in men (7.3%; 95% CI, 6.4–8.2). FI occurs at least weekly in 1.13% of population. FI consisted of liquid stool in the majority of cases. Table 2 shows prevalence of FI by composition and frequency of leakage per week. Incontinence by more than one consistency was reported by 28% (95% CI, 26.5–29.3) of participants with FI.
Table 2Prevalence of FI by Stool Type and Frequency of Leakage
Composition of leakage
Frequency of leakage (no. of bowel movements per wk)
NOTE. Groups of subjects with different composition of leakage are not mutually exclusive. All values are expressed as percents.
Figure 1 shows the prevalence of FI by age group and gender. There was a linear upward trend in the overall prevalence of FI by age group (P trend < .001) and sex. FI increased from 2.91% (95% CI, 2.22–3.81) in the 20- to 29-year-old subjects through 8.54% (95% CI, 7.38–9.87) in the 50- to 60-year-old subjects to 16.16% (95% CI, 14.15–18.39) in those 70 years and older.
Trends in Prevalence of Fecal Incontinence
The prevalence of FI did not change during the study period. The overall prevalence was 8.3 (95% CI, 7.2–9.5) in 2005–2006, 8.5% (95% CI, 7.5–9.6) in 2007–2008, and 8.4% (95% CI, 7.2–9.8) in the 2009–2010 survey. By gender, although the prevalence of FI decreased among men from 7.7% in 2005–2006 through 7.3% in 2007–2008 to 6.8% in 2009–2010, there was an upward trend among women from 8.9% in 2005–2006 through 9.6% in 2007–2008 to 10.0% in 2009–2010. Figure 2 shows the trends in FI prevalence from 2005–2010 by age group. The greatest increase in FI prevalence was observed in the 70 years and older group between the years 2007–2008 and 2009–2010. Figure 3 shows the trend in prevalence by gender and survey cycle. The prevalence of the different types of FI did not vary significantly during the 5 years.
Risk Factors Associated With Fecal Incontinence
Table 3 shows the bivariate and multivariate analysis for factors associated with FI. The significant limit for a factor to be included in the multivariate model was set at 0.1. From the multivariate analysis, risk factors for FI included age (>55 years), diabetes mellitus, male gender, UI, frequent and loose stools, poor health status, and ≥3 chronic illnesses. Obesity, parity, and number of pregnancies were not associated with FI. To further explore the reason for the switch from female to male gender (in the multivariate) as risk factor, the model was built forward by adding one variable at a time. We noted that this switch occurred when UI was introduced into the model. In addition, interactions between UI, age, and gender did not yield statistically significant effect modification. However, effect modification between age and gender was noted, with women being more likely to have FI with increasing age (odds ratio, 1.02; P = .03).
Table 3Bivariate and Multivariate Analysis for Factors Associated With FI
OR (95% CI)
Age group (y)
Body mass index (kg/m2)
Self-reported health status
Poverty index ratio
Frequency of bowel movements (per wk)
Vaginal deliveries (women only)
Chronic illnesses Baseline (none)
1 or 2
NOTE. Only variables found to be significant in age-adjusted bivariate analyses with alpha threshold of 0.1 were included in the multivariate regression and shown in this table. Multivariate odds ratios are adjusted for all other risk factors in the table.
n/a, not included in the multivariable logistic model.
Estimates on the prevalence of FI have varied widely. These inconsistencies have been mostly attributed to differences in sample populations and definitions of FI. Estimates from referral or specific subgroups of the populations overestimate the true magnitude of FI. NHANES is a national population-based survey of U.S. residents. The first national prevalence of FI among U.S. adults was reported in 2009.
However, this study used data only from the NHANES 2005–2006 cycle, thus it is limited in numbers. Our study used data during 3 survey cycles and represents the largest sample on FI at the population level in the United States. Estimates from this data set confirm previous findings from smaller studies conducted in the community around the United States.
However, this study included both institutionalized and non-institutionalized individuals, whereas NHANES includes only non-institutionalized individuals. FI is one of the primary reasons for admittance to nursing homes, and higher rates have previously been reported.
We believe that the true prevalence of FI lies somewhere between the numbers reported in community-based studies and those in long-term care centers.
This study documents trends in the prevalence of FI in the United States. We hypothesized that with the projected demographic changes in age in the United States, the prevalence of FI was on the rise. Our study showed that the prevalence of FI was relatively stable between 2005 and 2010. However, the prevalence increased steadily among the 70 years and older age group during the study period, as shown in Figure 2. The largest increase (3.0%, P < .001) in this age group occurred between 2007–2008 and 2009–2010.
Most discussions on the etiology of FI have been based on the assumption that women are more at risk than men. Yet, several population-based studies have reported no gender differences
In this study, we noted a male predominance in the younger age groups, with a gradual transition to female predominance in the older age groups. From 2005 to 2010, the prevalence decreased among men, whereas there was a steady increase among women. The reasons for these contrasting trends are unclear. It has been speculated that obstetric injuries probably contribute to the higher prevalence of FI among women.
This study showed that number of vaginal deliveries or parity was not associated with FI. Unfortunately, our data set did not have information to specifically assess the role of operative deliveries or obstetric injury.
One of our objectives was to identify risk factors for FI. Increasing age has been consistently shown to be associated with FI.
Some authors have argued that this association is probably confounded by limited mobility, multiple comorbidities, or overall poor health status. We were able to adjust for most of these factors, and yet age remained associated with FI. FI preventive strategies should focus on proper management of diabetes mellitus and control of diarrhea. There have been studies showing that when FI exists in the context of diarrhea, treating the latter often leads to improvement of the former.
rather than either being a risk factor for the other. We found an interesting relationship between FI, UI, and gender. In the age-adjusted bivariate analysis, female gender was strongly associated with FI. Surprisingly, in the multivariate analysis, men appeared to be at higher risk of FI. This switch was entirely explained by the presence of UI. This finding was not explained by effect modification between UI and gender or age. However, effect modification was noted, with women being more likely to have FI with increasing age. This result suggests that all risk factor analysis studies on FI should take into consideration the presence or absence of UI.
Recent reports have found a higher prevalence of FI among obese individuals.
Obesity might contribute to FI by increasing the intra-abdominal pressures. In this study, obesity (body mass index >30 kg/m2) was not associated with FI in a multivariate model. The association between obesity and FI is controversial. Whereas some authors have noted an improvement in FI with weight loss,
This study has several strengths. The majority of previous studies on FI have come from single institutions and were plagued by referral bias. Our sample is nationally representative, with all segments of the population including minorities well-represented. The standardization of the NHANES survey procedure and definitions of FI ensures high quality data collection. The NHANES survey design also partially bypasses the drawback of previous studies, which is the underreporting of FI owing to patients' reluctance to voluntarily report this to care providers.
Another strength of our study is the large number of subjects. The latter provides statistical power to study factors associated with FI. A few limitations of this study are worth mentioning. This analysis included only non-institutionalized individuals. These estimates are therefore an underestimation of the true magnitude of the problem because a significant proportion of individuals with FI live in nursing homes.
NHANES is a cross-sectional design, and thus incidence and causality cannot be established.
We have presented robust estimates of the prevalence, trends, and risk factors of FI in a nationally representative database of non-institutionalized adults in the United States. We found that FI is a common disorder in the United States, affecting about 19 million individuals as of April 2010. The prevalence of FI remained relatively stable between 2005 and 2010. Loss of liquid stool is the most common type of FI. Preventive strategies should focus on diabetes control and management of diarrhea. Future studies on the risk factors for FI should consider presence or absence of UI.
Supplementary Table 1Characteristics of Study Participants
Age group (y)
Poverty index ratio
Self-reported health status
Body mass index (kg/m2)
Frequency of bowel movements (per wk)
3 or more
NOTE. All results are reported as n (%). Respondents with missing data, “refused”, or responded with “don't know” were excluded from the table.