Prevention of urinary incontinence in older people


Fonda Director, Aged Care Services, Caulfield General Medical Centre, 260 Kooyong Road, Caulfield, Victoria 3162, Australia.


As health professionals, our goal is to cure incontinence. When this is not possible, we then seek to reduce the severity of the symptoms and failing that, to achieve a state of social continence with the use of aids and appliances [ 1]. Of course, our primary goal is preventing the onset of incontinence in the first instance [ 2]. There has been little work on the prevention of incontinence, so it is easier to illustrate the concepts of prevention using myocardial infarction as an example. Primary prevention would include, amongst other things, prevention or treatment of hypertension and hypercholesterolaemia to prevent coronary artery disease and consequent myocardial infarct. Secondary prevention would involve averting the occurrence of a second myocardial infarct, or the sequelae of the infarction, i.e. congestive heart failure. Tertiary prevention would involve treating the resultant congestive heart failure or reduced cardiac output.

 For the purposes of this paper, the following definitions of prevention will be used: (i) primary prevention; interventions that prevent the predisposing conditions (e.g. childbirth trauma, pelvic muscle weakness, detrusor overactivity, impaired mobility or mentation), from occurring: (ii) secondary prevention; to reverse the predisposing condition(s), or prevent their progression to incontinence: tertiary prevention; management strategies to decrease the severity and sequelae of incontinence.

 The principles of incontinence prevention in fit adults should apply equally to fit older people, although very few studies have determined whether treating identifiable risk factors translates into prevention of incontinence. The problem is further complicated when we add frailty and disability to this equation, as seen in many older people. In addition to frequent impairment of bladder innervation and sphincter function, such individuals often lack adequate mentation, motivation and mobility, and the likelihood of incontinence increases depending on how many of these functions are impaired. The clinical approach to each of these situations is somewhat different in the frail older group, compared with a fit young and healthy population. This review will draw on the available literature to present a conceptual framework for addressing the prevention of incontinence in this population.

The prevalence of urinary incontinence in older people

Estimates of the prevalence of urinary incontinence in this population varies widely. This is related to the absence of standardized and well-validated definitions and measures. The natural history of incontinence in this population is not well understood, but it is known that patient’s self-reporting of incontinence varies substantially even over a short period of time [ 3]. Interpretation of data is also difficult because levels of dependency differ in different residential settings and in different countries.

 The prevalence of urinary incontinence increases with increasing age and frailty. Among fit older people dwelling in the community, it ranges from about 5% in Singaporean women aged ≥65 years to about 30% in 85-year-olds living in Sweden [ 4, 5]. Among home-bound older adults, it is 13–53% [ 5[6][7][8]–9]. The incidence increases with age and it is more common in women [ 5]. It affects 25–35% of older patients admitted to acute hospitals [ 1, 10[11][12]–13].

 The prevalence of incontinence in nursing homes is even higher, ranging from 39% to 70%, with variability depending on admission criteria, definitions used and variability among countries [ 1, 14]. If the definition is extended to include ‘dependent continence’, i.e. dry only because of the assistance of staff or carers, then the value in nursing homes rises by at least a further 10% [ 1].

Difficulties with research in this population

There has been little work published on interventions to prevent rather than treat urinary incontinence in the older population. What work has been done has tried to identify risk factors or associated conditions that may be amenable to treatment. However, control data are lacking for the actual benefit of such outcomes. Some of the reasons for these difficulties are: (i) the heterogeneity of this population, resulting in difficulty in designing studies that account for comorbidity, drug use, intercurrent illness, and shorter life expectancy; (ii) a lack of noninvasive instruments to reliably classify the type of incontinence (e.g. stress, urge); (iii) a lack of information on the natural history of incontinence in this group; (iv) a lack of standardized terminology to define and measure cure and improvement; (v) a lack of validated research tools to measure baseline and outcome variables in the frail elderly; (vi) a lack of long-term follow-up to gauge the impact, durability and applicability of the intervention

 There are age-specific influences on lower urinary tract function but normative data are generally lacking in the older population. In addition, the test-retest reliability and sensitivity to change of the more invasive measures of lower urinary tract function are poorly documented in an older population, and especially in the frail and disabled elderly.

Risk factors for incontinence

Although there is an increasing awareness of risk factors associated with incontinence in older people, there are few data available on successful primary or secondary prevention strategies. Table 1 summarizes some of the known risk factors [ 15[16][17][18][19][20][21][22][23][24][25][26][27][28][29]–30]. However, identifying these factors does not establish a cause-and-effect relationship, and there are no data from controlled clinical trials that show interventions directed at these actually reduce incontinence incidents, severity or prevalence. Nonetheless, by concentrating on these risk factors, one hopes to reduce the prevalence or severity of the condition.

Table 1.  Risk factors associated with incontinence in older people Thumbnail image of

 Resnick et al. [ 15] recently performed a qualitative synthesis of 270 studies that sought risk factors for incontinence in older people. Because only one epidemiological study of the elderly assessed urinary incontinence risk factors longitudinally, information on secondary prevention was derived predominantly from cross-sectional studies that sought to identify factors associated with urinary incontinence. Although many of the studies included relatively few subjects with the putative risk factors, available data suggest that incontinence in the elderly is associated with age, female sex, impaired mobility, functional disability, cerebrovascular disease, diabetes mellitus, faecal incontinence, chronic cough, and various other lower urinary tract symptoms (LUTS). They found consensus for the association of geriatric urinary incontinence with cognitive impairment, constipation, parity, use of various drugs, history of genitourinary surgery, obesity, and cardiovascular disease. The few available data on alcohol, coffee and tea consumption do not confirm a positive association with incontinence. No epidemiological data were found for the association of incontinence with race, ethnicity or intake of various foods [ 15].

 Frailty and disability adds a further dimension to understanding preventive strategies. The definition used in this paper is ‘any person over 65 years of age, who does not leave their place of residence without assistance of others, or a person with dementia, or a person who has been admitted to a long-term care facility’. The definition extends to older people admitted to hospital. These people often suffer from multiple medical conditions and/or disabilities, which results in them becoming homebound or institutionalized. Because they require the assistance of others to perform some or all of the more basic activities of daily living (ADLs), including bathing, dressing, toileting and ambulating, results from younger populations or from older people with no disabilities cannot necessarily be extrapolated to this population.

 Some of the additional bivariate risk factors for incontinence identified in frail people are immobility and chronic degenerative disease [ 31], impaired cognition [ 32, 33], diuretics [ 20], faecal impaction [ 34], low fluid intake [ 35] and environmental barriers [ 36]. Multivariate analysis identified immobility, Parkinson’s disease, dressing dependence, dementia, stroke and diabetes as statistically significant risk factors for incontinence in a nursing-home population, whilst the role of faecal impaction could not be assessed [ 37].

 Urinary retention is a frequent occurrence in hospitalized elderly and is frequently unrecognized. In one study, a third of hospitalized elderly patients had a residual urine volume of >50 mL [ 38]. Patients with a fractured hip are particularly at risk of becoming incontinent, with over half having urinary retention in the postoperative period [ 39]. Stroke also carries a high risk of incontinence, with nearly half the patients being incontinent in the first week; the prevalence declines progressively thereafter [ 40]. In one urodynamic study, 47% of stroke patients had urinary retention [ 41].

 To assist in the identification and management of incontinence in nursing homes, a Minimum Data Set has been developed for nursing homes in the USA, which aims, amongst other things, to detect urinary incontinence. When this is present, the Resident Assessment Protocol can be applied, which provides a stepwise approach to managing continence [ 42]. Whilst the tool has been shown to be reliable in identifying incontinence, its impact on reducing incontinence rates remains to be determined.

Primary prevention

Hopefully, with the development of general principles for primary prevention in a younger, fitter population, there may be fewer older people who reach old age with predisposing conditions for incontinence. Improved obstetric management in developing countries should lead to a reduction in vesicovaginal fistula, the commonest cause of incontinence in this group of women. Alternatively, it may be shown that in the future, long-term or even lifelong pelvic floor exercises reduce the prevalence of pelvic floor weakness and incontinence in later life. Similarly, it may be shown that improved management of vaginal delivery or bowel-management protocols through adult and later life may reduce pudendal nerve damage and its sequelae in later life; or that prophylactic use of certain medications reduce prostatic size, thus reducing the predisposition for overflow incontinence and its sequelae in an older population. Measures that could reduce the prevalence of other medical conditions such as dementia, stroke and multiple sclerosis, might in turn reduce the incidence of incontinence by direct effects on the neural control of bladder functioning or secondarily by their effect on planning, memory or mobility.

Secondary prevention

The difficulty with secondary prevention of incontinence is that patients rarely present whilst they are still asymptomatic. Research to characterize those at higher risk is needed, as are longitudinal data to determine the role of various secondary preventive interventions directed at such persons. For example, will doing pelvic muscle exercises in a 50-year-old woman, who has a weak pelvic floor from childbirth, prevent incontinence?

 By the time a population has become frail and disabled, primary prevention is no longer achievable because of the existence of detrusor overactivity, pelvic muscle weakness, decreased mental status and/or impaired mobility. However, it is feasible to find secondary prevention strategies that are applicable to this population. The physical environment of the person, be it in their home, nursing home, or in the community at large, can significantly affect continence status [ 37]. Access to toilets, height of furniture and distance to toilets are some examples of this, as are any measures that might promote the physical independence of the person, or maximize their potential to be able to independently get to the toilet before being wet. This may be in the form of fitness and activity programmes, or rehabilitation programmes. Whilst such programmes seem to have improved fitness and reduced the incidence of falls, there are no data to show that they have reduced incontinence.

 Numerous drugs have been implicated in causing or contributing to incontinence and if these could be reduced or eliminated, the patient may avoid the indignity of iatrogenic incontinence [ 43]. The combination of various factors, e.g. diuretics or hypnosedatives, in an older person with impaired mobility is likely to lead to incontinence. Secondary preventive measures directed at their medication and mobility may cure the incontinence. Ultimately, all of the above preventive measures require a high level of community awareness, public education and health professional education.

 Besides the factors listed above, there is a high incidence of frequently unrecognized urinary retention occurring after hip fracture, and to a lesser extent stroke, which would increase the likelihood of persisting incontinence. This overflow incontinence is related to a combination of medical, cognitive, functional and environmental factors. Secondary preventive measures should identify and address the risk factors for the development of overflow incontinence in such patients before incontinence occurs. Unfortunately, as is the case in so many areas related to the frail, there are no available, reliable data that actually show that attention to such issues will lead to a significant reduction in urinary incontinence.

Tertiary prevention

In this context it is assumed that incontinence is present with or without bladder and/or related pathology, and measures are undertaken to cure or stop the progression of these symptoms. To address tertiary prevention equates to reviewing the literature of interventions to manage incontinence. Difficulties related to research design and lack of use of validated outcome measures have resulted in a paucity of good data, especially in the frail elderly (see below).

 Various treatment strategies used in an adult population should apply to a fit older population. However, the role of conservative options such as pelvic floor exercises, surgical interventions (other than to relieve obstruction) or environmental modification have not been shown to be of benefit in well-designed research studies in a frail population. There are no controlled studies showing that drug treatment has a significant role to play in the management of incontinence in the frail population [ 44].

 A range of behavioural management strategies has been tried in this frail older population, with variable success. Invariably, any benefit seen is dependent on the active, continuing cooperation and participation of nursing staff. This has been shown to be most difficult to sustain and invariably adds cost to the care in a nursing-home setting [ 45]. The reported short-term benefits of these behavioural interventions have been summarized recently in several publications [ 46[47]–48]. Ultimately, tertiary prevention measures in such settings can achieve dependent continence by use of regular toileting programmes, or social continence by use of various aids, including pads and/or catheters to contain the problem [ 1].

What needs to be done?

Given the current difficulties encountered with both managing and studying this particular population, there is a need to ‘get back to basics’. Because comorbidity and drug use contribute to the presence and severity of incontinence in this population, they should be addressed and/or stabilized before the patient is subjected to interventions of dubious benefit. The first step is to develop a sound basis for data collection by developing validated outcome measures that have been developed and tested in this population. There is a need to identify potentially reversible or modifiable subsets of patients who can be studied for various interventions. Ultimately, all of the above preventive measures require a high level of community awareness, public education and health professional education.

 If we are to progress in the prevention and management of incontinence in frail and older people, research methodology needs to improve, and use of validated outcome measures is essential. Further information has been provided on general guidelines for study design in incontinence by the ICS [ 49, 50]. These difficulties are well summarized in the position paper from the ICS Committee on Outcome Measures [ 49]. The world literature has been examined for validated tools that address patient’s symptomatology; objective and subjective measures (e.g. bladder diary, pad weighing); physiological measures (e.g. urodynamics, residual urine estimation); quality-of-life measures; and socio-economic measures. Of all of these measures, the only tools that appear validated for this population have been ‘wet checks’ [ 51, 52] and pad weighing [ 53]. With both strategies, this has required nursing staff to perform these measures, rather than the patient or their carer in their home situation. Therefore, if only validated outcome measurement tools were used, the number of research articles that could be reviewed, where bladder programmes are being instituted in this population, would be severely limited.

 The needs of a frail and/or disabled older population are different from both younger patients and healthy elderly patients, and study design and interventions should reflect this difference, especially the importance of defining clinical rather than statistical significance. We need to recognize that we should strive to make the patient comfortable (and possibly therefore dependent-continent) before we subject the patient to excessive and burdensome unproved interventions.


Although numerous factors are associated with incontinence in the community-dwelling elderly, many of the studies included few subjects with relevant conditions, consensus exists on only a few risk factors, temporal relationships have not been ascertained, type-specific data are lacking, and many of the factors such as age and sex are not amenable to intervention. Further research on the aetiology of the dysfunction that underlies incontinence, especially detrusor overactivity, is warranted. Targeted large-scale longitudinal studies are needed to define the group of older people at risk of becoming incontinent, as well as to devise and evaluate strategies for secondary prevention.

 Controlled longitudinal interventional studies of continent older adults are necessary to determine risk factors for incontinence in older people. In the community-dwelling elderly, whilst several risk factors have been identified in epidemiological studies, none has ever been proved by treatment. This is problematic, because association does not prove causality. Furthermore, risk factors in epidemiological studies may not equate to risk factors in individuals, especially the frail elderly who have been the subject of very little research on risk factors.

 In the absence of more definitive data, it is reasonable and appropriate to address each of the above mentioned ‘at risk’ characteristics in an attempt to avoid or at least postpone the onset of incontinence and improve quality of life. Ultimately all of the above mentioned preventive measures require a high level of community awareness, public education and health professional education. Some of these conditions might be amenable to primary prevention; (e.g. reduction of stroke incidence), secondary prevention; (e.g. reduced medications, reducing constipation, improving mobility, attention to cough), and some to tertiary prevention (e.g. functional disability, impaired mobility).

 The lack of progress in primary and secondary prevention of incontinence in an adult population is in sharp contrast to the significant advances in primary and secondary prevention in heart disease. Tertiary prevention (i.e. the treatment of the condition to cure or improve it, or prevent its deterioration), has been the major focus of clinicians to date. In the elderly, the opportunity for secondary prevention actually exists by attention to the various comorbid conditions, although data on secondary prevention suffers largely from methodological problems.


N.M.R. was supported in part by the NIA Teaching Nursing Home grant AG04390, NIDDK grant R01-DK49482, and Claude Pepper Older American Independence Centre grant AG-08812.