Epidemiology of fecal incontinence
Abstract
Nursing home residence is by far the most prominent association with fecal incontinence, with a prevalence approaching 50%. In one major survey, urinary incontinence was the greatest risk factor for developing fecal incontinence, and fecal incontinence was the greatest risk factor for developing urinary incontinence.
Immobility, dementia, and the use of physical restraints were also important risk factors. Specific diseases associated with fecal incontinence include diabetes, multiple sclerosis, Parkinson’s disease, stroke, and spinal cord injury. The surgical procedures lateral internal sphincterotomy for anal fissure, fistulotomy, and ileal pouch reconstruction can result in fecal incontinence. Children who are born with congenital abnormalities, such as imperforate anus, often experience soiling for many years.
Future studies to determine the prevalence and etiology of fecal and urinary incontinence will need to first define these conditions and eliminate referral bias. Epidemiologic investigations of both disorders should be performed jointly because the conditions are so often comorbid.
Article Outline
Most discussions of the etiology of fecal incontinence have been based on the assumption that women, particularly those younger than 65 years of age, are more at risk for fecal incontinence than men. Injury to the pudendal nerve or sphincter muscle from prior obstetric trauma is described as the primary risk factor, followed by irritable bowel syndrome (a disorder thought to be more prevalent in women) and neurologic diseases such as diabetes (a distant third). Yet each population-based survey of the prevalence of fecal incontinence has shown a surprisingly high prevalence in males. Clearly, causative factors other than childbirth must be sought. In the most broadly based survey, factors that affect general health or physical capabilities, independent of age and gender, place individuals at greater risk for fecal incontinence than either age or gender.
Evidence regarding fecal incontinence in epidemiologic investigations comes principally from cohort and cross-sectional surveys in which risk can be calculated and expressed as odds ratios with 95% confidence intervals and for which the statistical significance of the associations can be assessed. Secondary risk factors arise from case series and insightful observation. Systematic reviews of epidemiologic observations have been difficult to perform because different studies usually adjust for different variables and individual patient data are needed to overcome this barrier to quantitative statistical amalgamation of studies. Good evidence, such as that supplied by randomized controlled clinical trials, does not exist for any risk factor related to fecal incontinence. In the Tables that accompany this article, when odds ratios and confidence intervals are presented, it can be assumed that fair epidemiologic evidence exists in support of that risk factor. When only lists of associations are presented, the evidence is merely anecdotal.
Prevalence of fecal incontinence
Older reports of the prevalence of fecal incontinence have come from single institutions, and the patients described therein have been subject to referral bias when demographics and etiology are discussed. The accuracy of prevalence estimates for fecal incontinence may also be diminished by difficulty in ascertaining those figures and the common underreporting of fecal incontinence owing to patients’ reluctance to report symptoms or seek treatment (Table 1). It has been shown that women are more willing to report fecal incontinence than men. In addition, the character (incontinence of solid feces, diarrhea, or flatus) and frequency (daily versus episodic) of fecal incontinence vary greatly in each population.
Table 1. Prevalence of Fecal or Double (Fecal and Urinary) Incontinence in Middlesex, United Kingdom
Prevalence depends on the definition of fecal incontinence. The entry question quoted in one survey was written by a patient with fecal incontinence who regarded the inclusion of the symptom of gas as necessary. This individual and many others have stated that incontinence or incontinence only to solid stool need not occur frequently to make one believe that it is a constant, disabling condition. The most prominent risk factors for fecal incontinence in this and several other surveys were physical disability and poor general health (Table 2). Many population-based and cross-sectional studies of fecal and urinary incontinence in both community-dwelling and institutionalized individuals have concluded that prevalence varies depending on gender, age, health status, and place of residence (Table 3).
Table 2. Adjusted Odds Ratios for Fecal Incontinence Risk Factors
Table 3. Population-Based Surveys of the Prevalence of Incontinence
The most prominent association with fecal incontinence by far is nursing home residence. Whereas the prevalence of fecal incontinence is probably around 2% to 3% for community-dwelling persons and may increase with increasing age to greater than 10%, among nursing home residents the prevalence approaches 50%. Moreover, fecal incontinence is one of the most common reasons for nursing home admission. In a survey of 18,000 Wisconsin nursing home residents, risk factors for fecal incontinence were directly observed by nursing home personnel. Surprisingly, in this very old population (mean age, 84 years), age, gender, and diabetes were not found to be associated with fecal incontinence. Urinary incontinence was the greatest risk factor for fecal incontinence (and fecal incontinence was the most prominent risk factor for urinary incontinence), followed in order by the loss of ability to perform daily living activities, tube feeding, physical restraints, diarrhea, dementia, impaired vision, constipation, and fecal impaction. Inverse associations were noted with body weight, heart disease, arthritis, and (surprisingly) depression.
Pregnancy, although not the exclusive cause of fecal incontinence, is certainly a prominent association. Factors leading to incontinence during pregnancy, immediately after pregnancy, and long after pregnancy have been investigated. Irritable bowel syndrome has been shown to be an important correlate with postpartum fecal incontinence. Quantitative assessments of risk related to pregnancy and to various methods of delivery have only recently been performed. Interestingly, the risks associated with cesarean section and vaginal delivery are similar. However, in one study, the odds ratios were not adjusted for age or parity, and it is not known which cesarean sections were done emergently vs. electively (Table 4). Other reports suggest that some of the risk of fecal incontinence is attributable to having emergency sections.
Table 4. Association of Pregnancy and Method of Delivery With Incontinence to Feces and Flatus
Several specific diseases have been associated with fecal incontinence, and mechanisms to explain the associations have been investigated. These include diabetes, multiple sclerosis, Parkinson’s disease, spinal cord injury, systemic sclerosis, myotonic dystrophy, and amyloidosis (Table 5). Many of these conditions directly affect mobility and ability to perform daily living activities, or they cause diarrhea or fecal impaction. Children with congenital anal anomalies, such as imperforate anus, often have lifelong problems with incomplete evacuation and soiling despite anatomical correction. Other children are born without anomalies but - for various reasons - withhold stool at an age beyond which toilet training should be complete and develop fecal soiling or have megarectum. Failure to retrain the child at an early age often leads to chronic impaction and fecal incontinence.
Table 5. Diseases Associated With Fecal Incontinence
The importance of diarrhea in fecal incontinence cannot be overemphasized. One case series noted that 51% of individuals with chronic diarrhea were incontinent. In the Wisconsin Family Health Survey of fecal incontinence, 41% of the 25 subjects with fecal incontinence lived in Milwaukee at a time (April-May 1993) when the city experienced an outbreak of cryptosporidia in its drinking water supply, reportedly the largest outbreak of water-borne disease in United States history. This is an important reminder that infectious sources of incontinence should be part of the diagnostic evaluation of fecal incontinence when diarrhea is present. Noninfectious causes of diarrhea must also be considered, including leisure activities such as running. Other etiologies for fecal incontinence include stroke and hospitalization for acute illness.
Surgeons are often concerned about the possibility of fecal incontinence originating from surgery. This factor would seem fairly insignificant in the general population because prior anal surgery has not been an apparent risk factor in the larger surveys. Although surgical training is directed toward avoiding this disabling complication, several operations, nonetheless, frequently may result in minor fecal incontinence (Table 6). The first of these is lateral internal sphincterotomy for anal fissures. The risk of this procedure causing fecal incontinence was previously thought to be insignificant when compared with the risk associated with midline sphincterotomy (the most frequent procedure for anal fistula), but a recent reappraisal of this operation has shown a risk for fecal incontinence as high as 8%. Similarly, fistulotomy was thought to entail a negligible risk for fecal incontinence when compared with fistulectomy. However, recent estimates of the risk of fecal incontinence after fistulotomy have ranged from 18% to 52%. New approaches to fissure and fistula have recently been developed specifically to lower this risk. Ileal pouch anal reconstruction has enabled persons with inflammatory bowel disease to live without a stoma, but they remain at high risk for developing fecal incontinence. A more proximal anal anastomosis is now commonly performed with the hope of diminishing this risk. Should a pouch also be made for patients undergoing ileorectal anastomosis? Low anterior resection has made it possible for patients with midrectal cancer to avoid a permanent stoma, but the functional results - even in the absence of prior radiation - may be poor, and new procedures to improve outcomes have also been described. Last, mixing urine and stool has been found to have a predictable effect on anal sphincter control, as does diarrhea, in patients who undergo ureterosigmoidostomy after urinary bladder resection.
Table 6. Operations Associated With Risk of Fecal Incontinence
The development of incontinence in previously continent nursing home residents has also been studied. Significant associations with dementia, stroke, and blindness have been noted. However, the most significant association is with the use of patient restraints, even when adjustment has been made for factors that might be associated with restraint, such as immobility and dementia.
Fecal and urinary incontinence commonly coexist, particularly in the elderly and in nursing home patients. The prevalence of fecal incontinence increases with age, but the disorder is present in all age groups and both genders, varying from 1.5% in children to more than 50% in nursing home residents. Fecal incontinence is almost as common in men as in women. As populations age, comorbid disease becomes a significant component of incontinence risk. Epidemiologic investigations of fecal and urinary incontinence should be performed jointly. To reduce incontinence among nursing home residents, the use of truncal restraints in nursing homes should be reassessed.
References
1. K.A. Small and J.M. Wynne, Evaluating the pelvic floor in obstetric patients. Aust N Z Obstet Gynecol 30 (1990), pp. 41–45.
2. R.D. Madoff, J.G. Williams and P.F. Caushaj, Current concepts: fecal incontinence. N Engl J Med 326 (1992), pp. 1002–1007.
3. C. Abou-Zahr, Obstructed Labour. In: C.J.L. Murray and A.D. Lopez, Editors, Health dimensions of sex and reproduction: The global burden of sexually transmitted diseases, HIV, maternal conditions, perinatal disorders and congenital anomalies, Harvard University Press, Cambridge, MA (1996).
4. D.A. Drossman, What can be done to control incontinence associated with the irritable bowel syndrome?. Am J Gastroenterol 84 (1989), pp. 355–357.
5. L.R. Schiller, C.A. Santa Ana, A.C. Schmulen, R.S. Hendler, W.V. Harford and J.S. Fordtran, Pathogenesis of fecal incontinence in diabetes mellitus. N Engl J Med 307 (1982), pp. 1666–1671.
6. R.J. Leigh and L.A. Turnberg, Faecal incontinence: the unvoiced symptom. Lancet 1 (1982), pp. 1349–1351.
7. T.M. Thomas, M. Egan, A. Walgrove and T.W. Meade, The prevalence of faecal and double incontinence. Comm Med 6 (1984), pp. 216–220.
8. A.J. Campbell, J. Reinken and L. McCosh, Incontinence in the elderly: prevalence and prognosis. Age Ageing 14 (1985), pp. 65–70.
9. A.L.M. Kok, F.J. Voorhorst, C.W. Burger, P. van Houten, P. Kenemans and J. Jansens, Urinary and faecal incontinence in community-residing elderly women. Age Ageing 21 (1992), pp. 211–215.
10. P. Denis, E. Bercoff, M.F. Bizien, P. Brocker, P. Chassagne, H. Lamouliatte, A.M. Leroi, M. Perrigot and J. Weber, Prevalence of anal incontinence in adults [French] . Gastroenterol Clin Biol 16 (1992), pp. 344–350.
11. D.A. Drossman, Z. Li, E. Andruzzi, R.D. Temple, N.J. Talley, W.G. Thompson, W.E. Whitehead, J. Janssens, P. Funch-Jensen, E. Corazziari, J.E. Richter and G.G. Koch, US householder survey of functional gastrointestinal disorders. Prevalence, sociodemography, and health impact. Dig Dis Sci 38 (1993), pp. 1569–1580.
12. R.L. Nelson, N. Norton, E. Cautley and S. Furner, Community-based prevalence of anal incontinence. JAMA 274 (1995), pp. 559–562.
13. N.J. Talley, E.A. O’Keefe, A.R. Zinsmeister and L.J. Melton, Prevalence of gastrointestinal symptoms in the elderly: a population based study. Gastroenterology 102 (1992), pp. 895–901.
14. A.H. MacLennan, A.W. Taylor, D.H. Wilson and D. Wilson, The prevalence of pelvic floor disorders and their relationship to gender, age, parity, and mode of delivery. BJOG 107 (2000), pp. 1460–1470.
15. N. Nakanishi, K. Tatara, K. Nakajima, H. Takabayashi, S. Takahashi, H. Naramura and K. Ikeda, Urinary and fecal incontinence in a community-residing elderly population: prevalence, correlates, and prognosis [Japanese] . Nippon Koshu Eisei Zasshi 44 (1997), pp. 192–200.
16. G.D. Giebel, R. Lefering, H. Troidl and H. Blochl, Prevalence of fecal incontinence: what can be expected?. Int J Colorectal Dis 13 (1998), pp. 73–77.
Richard L. Nelson
* Department of Surgery, University of Illinois College of Medicine at Chicago and Epidemiology/Biometry Division, University of Illinois School of Public Health, Chicago, Illinois, USA