Fecal incontinence is an underreported and underappreciated problem in older adults. Although fecal incontinence is more common in women than in men, this difference narrows with aging. Risk factors that lead to the development of fecal incontinence include dementia, physical disability, and fecal impaction. Treatment options include medical or conservative therapy for older adults who have mild incontinence, and surgical options can be explored in selected older adults if surgical expertise is available.

Syed H. Tariq MD, FACP

Department of Internal Medicine, Divisions of Geriatric Medicine, Saint Louis University School of Medicine, 1402 South Grand Avenue, M-238, Saint Louis, MO 63104, USA

Prevalence and Correlates of Fecal Incontinence in Community-Dwelling Older Adults

Fecal incontinence is a condition in which social stigma plays a major role in patient’s lives, both in reporting of the problem and in quality of life.^[1–3] Fecal incontinence can have tremendous negative effect on a person’s social interactions and self-image and can lead to restricted living space and poorer self-perceived health.^[4–6] The failure of physicians to inquire about fecal incontinence and the embarrassment surrounding the condition can deter patients from reporting it and from receiving treatment.^[2,4] The condition is largely regarded as an inability to care for oneself and is the second most-common cause for nursing home admission in the United States.^[2,4,7] Fecal incontinence is costly to the patient and to society in terms of lost nursing time spent in cleaning up after accidents, protective pads, and clothing, and laundering of soiled bed linens. ^[4,7–9] Many individuals have reported that fecal incontinence limits their social activities and may result in an inability to work or travel. Although fecal incontinence has a major effect on society and on individuals, research on the topic tends to be scarce and variable.

Reports of the prevalence of fecal incontinence from population-based studies of community-dwelling older adults range from 3.0% to 16.9%.^{[2,3,5,8,10,11]} One of the problems in comparing studies of fecal incontinence is the variability in the definition of the condition, particularly in terms of which level of severity constitutes incontinence. For example, one study of adults aged 40 and older^[1] found a 1.4% prevalence of major fecal incontinence, defined as “soiling of underwear or worse with a frequency of several times a month or more,” whereas another study^[12] found a 17% prevalence in adults aged 70 and older when asking the question, “In the past few months, have you ever lost control of your bowels when you didn’t want to?” The effect of different definitions can also be seen in two studies of older adults living in the same county. In studying those aged 50 and older, one found prevalences of 11.1% in men and 15.2% in women when asked whether they had had stool leakage in the previous year.^[8] Another asked about stool leakage more than once a week in participants aged 65 and older and reported a prevalence of 3.1% in women and 4.5% in men.^[3]

Previous research on risk factors for fecal incontinence have identified several factors: advancing age, diabetes mellitus, urinary incontinence, stroke, physical limitations, female sex, peri-anal injury or surgery, gynecological surgery, hypertension, poor general health, and bowel-related factors such as feelings of incomplete defecation, constipation, straining at stool, fecal urgency, and loose or watery stools.^{[4–6,8,10,13,14]} Although female sex is given by many as a risk factor, some studies have found that fecal incontinence is as much a problem in men as it is in women.^[1,4,8] In one study, fecal incontinence was associated with advancing age in men but not in women.^[8]

The present study reports a population-based study of adults aged 65 and older in Alabama, stratified by sex, race (white vs African-American), and rural versus urban dwelling. The purposes of the study were to estimate the prevalence of fecal incontinence in this community-based sample and to identify risk factors for the presence and severity of fecal incontinence.

Methods

From November 1999 to February 2001, the University of Alabama at Birmingham (UAB) Study of Aging enrolled 1,000 participants from the Medicare beneficiary lists of five counties in west-central Alabama. The lists were stratified by sex, race, and urban versus rural. Three of the counties were classified as rural and two as urban. The five counties have a significant proportion of African-American residents (20% to 65%), and were chosen for their proximity to the study centers. The sample was specifically selected to include 25% black women, 25% black men, 25% white women, and 25% white men.

Participants selected from the Medicare lists received a letter about the study in the mail and then received a telephone call to arrange an in-home interview. Participants who were unable to understand the recruiter or failed to arrange an in-home interview were excluded from the study. After the interview, participants’ personal physicians were contacted by mail to confirm medical diagnoses. Hospital discharge reports were requested for participants who had been hospitalized within 3 years of the interview. The UAB institutional review board reviewed and approved the study protocol. All participants signed informed consent forms before the beginning of the interview.

Interviews lasted about 2 hours. Using a structured questionnaire, trained interviewers asked participants about sociodemographic factors, medical conditions, health behaviors, activities of daily living (ADLs), life-space mobility, and self-reported health status. ADLs included bathing or showering, dressing and undressing, eating, walking, getting out of bed or chair, getting outside, and getting to and using the toilet. ADL impairment scores ranged from 0 to 7, with higher scores indicating impairment on more ADLs. Instrumental activities of daily living (IADLs) included using the telephone, doing light housework such as washing dishes and dusting, doing heavy housework such as yard work or vacuuming, preparing meals, shopping, and managing money. IADL impairment scores were scored from 0 to 6, with higher scores indicating impairment on more IADLs. Life-space mobility was defined by the distance a person traveled to perform activities during the month before the interview; 0 life space was restricted to the bedroom, 1 to the home, 2 outside the home, 3 the neighborhood, and 4 the town, and 5 was unrestricted. Independent life-space was defined as the highest level of life space achieved without help from any equipment or other persons.^[15] Subjects were also asked how they would describe their physical activity level compared with that of their adult life before age 65: less active, about as active, more active. In addition, the following performance measures were assessed: time to walk 9 feet and time to stand up from a firm seat five times with arms folded across the chest. Scores of 0 to 4, with 4 indicating the best performance, were assigned according to the quartile of the participant’s time to complete the task.^[16,17] The 15-item Geriatric Depression Scale was used to evaluate depressive symptoms.^[18] A comorbidity index was created giving one point for each disease category of the Charlson Comorbidity Index^[19] without regard to severity.^[15]

The data were entered on a TELEform (1999) version of the questionnaire. After a manual inspection, paper forms were corrected for errors and completeness and then scanned. The TELEform computer scannable system was programmed to detect errors or missing data. Any missing or incorrect data were immediately identified, and participants were recontacted if necessary to obtain corrected data. SPSS 2001 was used for statistical analyses (SPSS, Inc., Chicago, IL).

Fecal incontinence was defined as an affirmative response to the question: “In the past year, have you had any loss of control of your bowels, even a small amount that stained the underwear?” To determine severity of incontinence, questions were asked about frequency of incontinence. Mild incontinence was defined as having less than one accident a month, and moderate/severe incontinence was defined as having one or more accidents a month. Variables that were believed to be potential risk factors for fecal incontinence were identified.

Univariate analyses were conducted to evaluate which factors were correlated with fecal incontinence, using chi-square analyses for categorical variables and analysis of variance for continuous variables. After this screening, variables with P <.15 were entered into a multivariable forward stepwise logistic regression. Variables that were highly correlated were entered separately and the resulting models evaluated for goodness of fit.

Results

The characteristics of the 1,000 participants are shown in Table 1 . By design, the sample consisted of 25% white women, 25% white men, 25% black women, and 25% black men. The prevalence of any fecal incontinence was 12%, 12.4% in men and 11.6% in women. Sex and race differences were not statistically significant ( P =.33;). The prevalence of moderate/severe fecal incontinence (one or more accidents/month) was 5.8% in men and 4.6% in women ( P =.40) and did not differ by race ( P =.57).

Variables that were significantly associated with fecal incontinence in the forward stepwise logistic regression analysis are shown in Table 4 . Factors associated with more frequent fecal incontinence, defined as fecal incontinence once a month or more, were similar to those for any fecal incontinence within the previous year. For women, chronic diarrhea (odds ratio (OR)=6.39, 95% confidence interval (CI)=2.25–18.14; P <.001), poor self-perceived health (OR=5.37, 95% CI=1.75–16.55; P =.003), and urinary incontinence (OR=4.96, 95% CI=1.41–17.43; P =.01) were associated with more frequent fecal incontinence. For men, chronic diarrhea (OR=5.38, 95% CI=1.77–16.30; P =.003), poor self-perceived health (OR 3.91, 95% CI=1.39–11.02; P =.01), lower extremity swelling (OR=2.86, 95% CI=1.20–6.81; P =.02), and increased assisted life-space mobility (OR=0.73, 95% CI=0.49–0.80; P =.02) were associated with more frequent fecal incontinence.

Discussion

The findings of this study confirm that fecal incontinence is quite common in community-dwelling older adults, most likely more common than is generally assumed by healthcare practitioners. A condition such as fecal incontinence that is of fairly high prevalence, disturbing to the patient, and quite treatable should be included in the review of systems of primary care providers and appropriate specialists. As has been found in other studies, the prevalence of fecal incontinence was similar in men and women.^[1,4,5] Only one community-based study found fecal incontinence slightly more common in women.^[8] There were no racial differences in prevalence of fecal incontinence between African Americans and whites. The power to detect racial differences due to oversampling of African Americans is a particular strength of this study, because no prior study has had sufficient power to examine racial differences in prevalence of fecal incontinence. This contrasts with the literature on racial differences in urinary incontinence, which shows a lower prevalence among African-American women.^[20] More epidemiological research is needed to further examine racial differences or lack of them in the prevalence of fecal incontinence.

      For both men and women, the strongest predictor of fecal incontinence was chronic diarrhea, with an odds ratio of 4.6 in women and 6.1 in men. More than one-quarter of men and women with fecal incontinence reported chronic diarrhea. The relationship is probably causative, because liquid stool is more difficult to control than solid stool. It is quite likely that even more patients with chronic diarrhea would have fecal incontinence, but prevent it by limiting their activities so that they stay close to a bathroom. The etiologies of chronic diarrhea are most likely multifactorial, including irritable bowel syndrome, inflammatory bowel disease, constipation with overflow, lactose intolerance, chronic       Clostridium difficile      , and postcholecystectomy diarrhea, among others. Once chronic diarrhea is uncovered on review of systems, a structured assessment is necessary to identify the cause of the diarrhea and then targeted treatment undertaken. Simple treatments such as a bowel regimen for chronic constipation or fiber supplements for irritable bowel syndrome can often control chronic diarrhea and eliminate fecal incontinence. Research on the outcome of an evaluation for chronic diarrhea, the specific etiologies, and the effect of their treatment on fecal incontinence is needed.     

Fecal incontinence was also associated with prostate disease in men but not with prostate surgery. In previous research, radical prostatectomy has been associated with fecal incontinence.^[21] The current study may have failed to detect such a relationship because it did not differentiate between surgery for benign prostatic hypertrophy and the generally more invasive surgery for cancer, radical prostatectomy. Pelvic surgery has also been described as a risk factor for fecal incontinence in women,^[14] consistent with the finding of the current study that hysterectomy was associated with increased prevalence of fecal incontinence. The association was significant for hysterectomy with oophorectomy but not for hysterectomy without oophorectomy. It is possible that hormonal status has a role in fecal incontinence,^[22] although no difference was found in prevalence of incontinence between women on hormone replacement therapy and those not taking hormones.

Fecal incontinence was found to be associated with urinary incontinence in women but not in men in the multivariate analysis. This correlation has been reported in men and women other studies.^[8,10] Urinary and fecal incontinence are believed to share etiological factors in women, some of which may include damage to the pelvic floor sustained through childbirth or surgery (hysterectomy).

In men, transient ischemic attack, or ministroke, was found to be significantly associated with incontinence, but major stroke was not. Studies have shown that patients often have urinary or fecal incontinence after an acute stroke, and that some experience long-term incontinence.^[23–25] It is possible that individuals who had had a stroke severe enough to cause fecal incontinence were no longer community-dwelling but resided in long-term care facilities, where they were not included in this study. It is also plausible that even patients with minor strokes have lasting impairment to continence, a possibility that deserves further study.

Some of the other correlates also support a causal role for fecal incontinence on quality of life. In men, for example, fecal incontinence was associated with higher scores on the Geriatric Depression Scale. Men with fecal incontinence were also more likely to live alone. It is highly plausible that fecal incontinence contributes to depression, as well as to perception of poor health. This deserves to be explored further, because treating a patient’s fecal incontinence may provide substantial improvement in the patient’s perception of health and quality of life.

One limitation of all studies of fecal incontinence is the lack of consensus on the definition. The definition chosen for this study was designed to be more inclusive of amount of stool lost, because in clinical practice, even seepage of small amounts of stool can cause odor and social isolation. The definition was restricted to occurrences within the previous year. Studies that define fecal incontinence as including one episode of incontinence ever may include patients who may have only had an episode during an acute diarrheal illness or during delirium experienced during a hospitalization. Although this may be meaningful, the study definition was targeted to include fecal incontinence more likely to require clinical evaluation. As a measure of severity, frequency of fecal incontinence was asked. An additional limitation of the study was that the definition was restricted to fecal incontinence versus anal incontinence and thus did not include incontinence to flatus, which can be quite distressing to some patients.^[20]

In conclusion, fecal incontinence is more common in community-dwelling older adults than is generally assumed and should be included in the review of systems for older persons, especially those with chronic diarrhea.

Patricia S. Goode; Kathryn L. Burgio; Anne D. Halli; Rebecca W. Jones; Holly E. Richter; David T. Redden; Patricia S. Baker; Richard M. Allman