Natural History and Effects on 2-Year Outcomes of Urinary Incontinence After Stroke
Background and Purpose—We sought to describe the natural history of poststroke incontinence and estimate its effect on survival and 2-year outcomes in stroke survivors.
Methods—Two hundred thirty-five incident cases of stroke in 1995 were classified by continence status at 10 days after stroke. Age, sex, ethnicity, diabetes, hypertension, atrial fibrillation, premorbid disability, and Oxfordshire Community Stroke Project classification were recorded. Outcome data collected at 3 months and at 1 and 2 years included disability, case-fatality rates, and institutionalization rates. Disability was classified as severe, moderate, mild, or independent using the Barthel Index (without its “continence” component: 0–9, 10–14, 15–17, and 18, respectively) and Frenchay Activity Index (0–15, 16–30, and 31–45).
Results—Of 235 cases, 95 were initially incontinent (group 1); 140 were continent (group 2). At the initial, 3-month, and 1- and 2-year assessments, incontinence was recorded in 95 patients (40%), 34 (19%), 23 (15%), and 12 (10%), respectively. In univariate analyses, the 2 groups were not different in terms of demographic factors and risk factors. Compared with group 2, group 1 patients were more likely to have atrial fibrillation (28% versus 16%; P=0.02). Multivariate analyses showed that age >75 years (OR 15.9; CI 2.2 to 116.2), dysphagia (OR 4.03; CI 1.85 to 8.73), motor weakness (OR 5.41; CI 1.38 to 21.1) and visual field defects (OR 4.78; CI 1.78 to 12.9) were all significantly associated with incontinence. Incontinence was less common in lacunar infarctions (OR 0.12; CI 0.02 to 0.62). At 2 years, compared with group 2, group 1 had higher case-fatality rates (67% versus 20%; P<0.001), higher institutionalization rates (39% versus 16%; P=0.007), and greater disability (Barthel [0–9]: 39% versus 5%; P<0.001; Frenchay [0–15]: 75% versus 37%; P=0.001). Death or disability at 2 years was worse in subjects with initial incontinence(OR 4.43; CI 1.76 to 11.2).
Conclusions—Incontinence remains a prevalent condition 2 years after stroke. Initial incontinence was associated with age >75 years, dysphagia, visual field defect, and motor weakness. Poststroke incontinence adversely affected 2-year stroke survival, disability, and institutionalization rates.
Many studies1 2 4 7 8 9 10 have also reported that urinary incontinence is a strong predictor of mortality and poor functional outcome of stroke. Taub and colleagues1 showed that initial incontinence in first-time stroke survivors aged <75 years was the best single predictor of severe or moderate disability at 3 months, with a sensitivity of 60% and specificity of 78%. Similarly, Wade and colleagues2 showed that incontinence had an overall predictive value of 78% for death and disability at 6 months. Previous studies have also shown that urinary incontinence is an important factor for stroke mortality.4 11 However, these studies did not distinguish between incontinence that began before or after onset of stroke, and one study4 looked only at hospital admissions, perhaps biasing their sample toward those more severely affected.
Little is known about the natural history of this impairment after stroke. Furthermore, only 1 small study, by Brocklehurst in 1985,3 has examined the effect of urinary incontinence on stroke outcome beyond 12 months.
The objectives of this population-based observational study were to describe the natural history of urinary incontinence after stroke over a 2-year period and to estimate the effects of initial incontinence on mortality and 2-year outcome after stroke.
Subjects and Methods
Stroke patients were recruited from the South London Stroke Register (SLSR), which is a prospective, population-based stroke register covering a population of 234 533. The detailed methodology of the register has been reported in an earlier study.12 Data are collected on first-in-a-lifetime strokes in subjects of all age groups.
Stroke is defined according to the WHO criteria.13 The diagnosis of stroke and the initial assessment are made by one of the study doctors within the first week after the event (as far as possible). The classification of the subtype of stroke is made on the basis of clinical examination, radiological (CT or MRI) findings, analysis of cerebrospinal fluid, or necropsy examination. The Oxfordshire Community Stroke Project (OCSP) Classification is used to distinguish between lacunar and nonlacunar infarcts.14 Follow-up data are collected by a combination of personal interviews with the patients or their carers and postal questionnaires. For this study, data were obtained for patients who had been followed up for 2 years. Hence, all patients who were registered from January 1, 1995, until December 31, 1995, were included in the study.
Patients who had incontinence before their stroke (premorbid incontinence) were excluded from the study. On the basis of their continence status at approximately 7 to 10 days after their stroke (the “initial” period), the remaining patients were divided into 2 groups: incontinent (group 1) and continent (group 2). Patients were considered incontinent if they had lost bladder control or had an indwelling catheter within 48 hours of assessment. The 7-day period was considered necessary to allow for stabilization of the medical condition; to eliminate the confounding influence of fluctuations in consciousness, impairment, and disability in the immediate poststroke period; and to reduce bias due to early mortality.
Data that were collected included age, sex, and ethnicity; past history of risk factors for stroke, such as smoking, alcohol, hypertension, diabetes, atrial fibrillation, or transient ischemic attack; indicators of initial stroke severity, including dysphagia, dysphasia, motor deficit, and visual field defect; pathological and OCSP classifications of stroke; and premorbid disability, determined by using the Barthel Index (BI).
When possible, patients were followed up at 3 months and at 1 and 2 years after stroke. At each of these stages, incontinence was defined as “losing bladder control at least once a week or having an indwelling catheter.” The outcome measures collected at each time point were survival rates; disability indexes, including the BI and Frenchay Activities Index (FAI)15 ; Rankin scale16 scores; and institutionalization rates. The BI was used because it has a well-established validity,17 communicability, and efficiency,18 and it is simple to use. Because urinary continence status is one of the components of the BI, we excluded this component to avoid any obvious interaction. Therefore, in our BI, the maximum score was 18. The BI was interpreted in accordance with previous studies as follows:2 19 0 to 9, severely disabled; 10 to 14, moderately disabled; 15 to 17, mildly disabled; and 18, functionally independent. These ranges are similar to those identified by Wolfe and colleagues19 as corresponding closest to the Rankin scale for handicap.16 The FAI was also used as it measures the instrumental activities of daily living, including complex activities such as hobbies, household, and recreation. The FAI was interpreted as follows: 0 to 15, inactive; 16 to 30, moderately active; and 31 to 45, very active.11 Handicap was measured with the Rankin scale.16 Data on the 2 groups were compared by examining the effect of initial incontinence on mortality and 2-year outcome in stroke survivors.
Statistical analyses were performed using the STATA statistical analysis package (Stata Corp). Univariate analyses were carried out by using the appropriate statistical tests: confidence intervals, χ2 for categorical data, and t test for continuous data. To identify factors that were independently associated with poststroke urinary incontinence, backward stepwise multiple logistic regression was performed. Factors entered into the model included those that were significant at the univariate level (atrial fibrillation, dysphasia, dysphagia, visual field defect, and motor weakness) and other factors that were considered clinically relevant (age groups [<50, 50 to 75, >75 years], sex, OCSP classification, diabetes, and hypertension). Model selection was based on the likelihood ratio statistic (Pe=0.05, Pr=0.1). Odds ratios, controlling for age and sex, were determined from the final model. Multiple logistic regression was used to determine the associations of clinically relevant factors on death or disability at 2 years, where disability was defined as BI score of 0 to 14. This model included age group, sex, OCSP classification, initial incontinence, motor weakness, dysphasia, dysphagia, and visual-field defect.
Two hundred ninety-four patients with first-in-a-lifetime strokes were registered during the study period, ie, between January 1 and December 31, 1995. Fifty-nine patients were excluded from the study: 16 had premorbid incontinence, 32 patients died before initial assessment, and 11 stroke patients were registered late so their initial assessments were incomplete. Thus, 235 patients (80%) were studied. Of the 16 subjects with premorbid incontinence, 3 died within first week of stroke, another 7 died within 3 months, and 4 subjects died within 2 years. Thus, only 2 of the 16 were alive at 2 years, 1 of whom was still incontinent.
Natural History of Incontinence
There were 95 patients (40%) who were incontinent on initial assessment (Table 1⇓). At 3 months, 34 of the stroke survivors (19%) had incontinence. One year after stroke, 23 of the stroke survivors (15%) had incontinence; after 2 years, 12 of the stroke survivors (10%) had incontinence. Prevalence rates of incontinence were also estimated.
The Figure⇓ describes the natural history of poststroke incontinence over 2 years. Of the 95 patients incontinent at 7 to 10 days, 19 were incontinent and 32 were continent at 3 months; of the 19 incontinent, 8 were incontinent at 1 year and 3 at 2 years. Of the 32 continent, 5 were incontinent at 1 year and 3 at 2 years.
Of the 140 patients continent at 7 days, 15 were incontinent and 114 were continent at 3 months; of the 15 incontinent, 4 were incontinent at 1 year and 3 at 2 years. Of the 114 continent, 6 were incontinent at 1 year and 3 at 2 years.
Characteristics of Patients
Table 2⇓ illustrates the baseline characteristics of the 235 patients. There were no differences between the patients who were incontinent (group 1) and continent (group 2) in terms of sociodemographic factors, including age, sex, and ethnicity. There were no statistically significant differences between these groups in terms of history of current smoking, alcohol, hypertension, diabetes, or transient ischemic attack. Atrial fibrillation was significantly more prevalent in the incontinent group of patients (27 [28%] versus 22 [16%; P=0.02]). Premorbid disability, measured with the BI, was comparable between the 2 groups. Classification of stroke according to infarction, intracerebral hemorrhage, and subarachnoid hemorrhage revealed no differences between the 2 groups. However, when stroke was classified by OCSP classification, incontinence was significantly more prevalent in patients with total anterior circulatory infarctions than in patients with lacunar or posterior circulatory infarctions (P<0.001).
Table 3⇓ shows the multivariate analysis of factors that were independently associated with initial urinary incontinence. Factors with significant associations were age >75 years, dysphagia, visual field defect, and motor weakness. Incontinence was significantly less associated with lacunar infarctions.
Influence of Incontinence on Long-Term Outcome
Table 4⇓ shows the comparisons between the outcomes of groups 1 and 2 at 3 months and at 1 and 2 years using univariate analyses. The case-fatality rates were significantly higher in the incontinent group at all 3 time points. Follow-up of survivors was achieved in 98%, 94%, and 86% at 3 months, 1 year, and 2 years, respectively. The institutionalization rates were higher in group 1 at all times. According to the BI, more subjects in group 1 were severely (39%) or moderately disabled (18%) at 2 years, whereas group 2 had more subjects who were mildly disabled (26%) or independent (51%). This pattern of disability is further confirmed by the FAI, which also revealed that the initially “incontinent” patients were more disabled than the “continent” patients at all 3 time points. At 2 years, handicap, measured using Rankin scale, was also worse in the incontinent group than in the continent group (P=0.013).
Table 5⇓ gives the odds ratios for several clinically relevant factors in predicting death or disability at 2 years with use of multiple logistic regression. Urinary incontinence had an OR of 4.4 in predicting this outcome measure. Motor weakness was also a good predictor of outcome (OR 3.7) at 2 years, but dysphasia, dysphagia, and visual field defect were not statistically significant in predicting outcome in these patients. Patients with lacunar (OR 0.07), posterior circulatory (OR 0.08), and partial anterior circulatory (OR 0.09) infarctions had better outcome at 2 years than did those with total anterior infarctions.
This study is the first population-based study to describe the natural history of urinary incontinence following stroke over 2 years, excluding premorbid incontinence, and to estimate the effects of poststroke incontinence on 2-year outcomes. We have determined the following prevalence rates of urinary incontinence up to 2 years after stroke: 40% at 7 to 10 days, 19% at 3 months, 15% at 1 year, and 10% at 2 years.
The association between stroke and incontinence in previous studies is confounded by the lack of precise differentiation between premorbid and stroke-related incontinence. A review of stroke and incontinence20 included only 1 study21 that differentiated between premorbid and stroke related incontinence. Jawad and colleagues22 reported a significant relationship between premorbid urinary incontinence and stroke functional outcome at 6 months. The prevalence rates of urinary incontinence in stroke patients in the Copenhagen Stroke Study4 on admission and at 6 months were 47% and 19%, respectively; however, this was a hospital-based study and thus may have missed the cases not admitted to the hospital, and it did not exclude subjects with premorbid incontinence. According to the study of Wade and Hewer,2 44% of 532 patients were incontinent in the first week of stroke and 11% after 6 months. Borrie et al21 found that incontinence in hospitalized stroke patients at 4 and 12 weeks was 42% and 29%, respectively. Brocklehurst and colleagues3 reported prevalence rates of 18% at 2 weeks and 12% at 6 months.
Multivariate analysis indicated that urinary incontinence was significantly associated with age >75 years, motor weakness, dysphagia, and visual field defect. Sex, diabetes, hypertension, atrial fibrillation, and dysphasia were not associated with initial urinary incontinence. We did not include loss of consciousness in our analysis of factors associated with incontinence because all subjects who were unconscious were, by definition, incontinent; they were either unable to control their bladder or they were catheterized. We used the OCSP classification of stroke14 in our study to determine whether incontinence was associated with any particular subtype of stroke. Patients with lacunar infarctions were less likely to suffer from poststroke urinary incontinence than those with total anterior circulatory infarctions. Because the actual numbers of subarachnoid hemorrhages were small, no conclusions could be drawn from the odds ratios associated with subarachnoid hemorrhages in the model.
This study has shown that urinary incontinence has an adverse relationship with stroke outcomes at 3 months, 1 year, and 2 years. The mortality and institutionalization rates were significantly worse in the incontinent group of patients at all 3 of these time points. Disability, measured with the BI and FAI, was also worse in this group. Handicap, measured with the Rankin scale, was worse in the incontinent group. Multivariate analysis has shown that urinary incontinence is a strong predictor of death or disability at 2 years. These results are contrary to those in the study of Brocklehurst et al,3 which reported that incontinent patients had no significant differences in dependency or hospitalization at 1 year compared with their whole series. However, because the numbers in that study were quite small, it was difficult to estimate long-term effects of incontinence on stroke outcomes.
Previous studies1 2 8 have reported adverse effects of incontinence on stroke outcomes that are similar to those in our study. However, these studies were restricted to short- and medium-term outcomes, and none of them differentiated between premorbid and poststroke incontinence. Studies by Wade and Hewer2 and Barer8 described the prevalence and effects of urinary incontinence at 6 months; the Barer study looked at hospital patients only. Taub and colleagues1 excluded patients over the age of 75 which, as shown in our study, is a significant factor for poststroke urinary incontinence.
Contrary to hospital-based studies, our study was population based and thus included patients who died out of the hospital as well as those with mild manifestation for whom referral for hospital rehabilitation may not have been considered. We excluded all patients with premorbid incontinence, thereby ensuring that all our results were based on poststroke incontinence. Our follow-up rates of stroke survivors were consistently high at all 3 time points. Hence, any differences in outcomes cannot be attributed to lack of an adequate number of follow-ups. This study could not assess the precise reasons for incontinence in each subject. Moreover, we could not determine whether any interventions were used on any of the subjects to manage incontinence. Lack of this information also meant that we could not estimate benefits (and effects on stroke outcomes) of any particular treatment modality for incontinence that may have been used on the subjects in the study (eg, the effects of using antimuscarinic agents for poststroke incontinence).
We conclude that poststroke urinary incontinence is a prevalent condition for up to 2 years after stroke. Initial urinary incontinence was associated in multivariate analysis with age >75 years, dysphagia, visual field defect, and motor weakness, but not with sex, diabetes, hypertension, atrial fibrillation, and dysphasia. Initial incontinence was significantly less common in subjects with lacunar infarctions. Two-year stroke survival, disability, handicap, and institutionalization rates were adversely influenced by poststroke incontinence.
Mehool Patel is a Clinical Research Fellow funded by the Stroke Association. The study was principally funded by the Northern and Yorkshire Region Cardiovascular Disease and Stroke Research and Development Programme.
- Received May 9, 2000.
- Revision received September 13, 2000.
- Accepted September 13, 2000.
- Copyright © 2001 by American Heart Association
Taub NA, Wolfe CD, Richardson E, Burney PG. Predicting the disability of first-time stroke sufferers at 1 year: 12-month follow-up of a population-based cohort in Southeast England. Stroke. 1994;25:352–357.
Wade DT, Hewer RL. Outlook after an acute stroke: urinary incontinence and loss of consciousness compared in 532 patients. Q J Med. 1985;56:601–608.
Nakayama H, Jorgensen HS, Pedersen PM, Raaschou HO, Olsen TS. Prevalence and risk factors of incontinence after stroke: the Copenhagen Stroke Study. Stroke. 1997;28:58–62.
Kalra L, Smith DH, Crome P. Stroke in patients aged over 75 years: outcome and predictors. Postgrad Med J. 1993;69:33–36.
Gladman JR, Harwood DM, Barer DH. Predicting the outcome of acute stroke: prospective evaluation of five multivariate models and comparison with simple methods. J Neurol Neurosurg Psychiatry. 1992;55:347–351.
Barer DH. Continence after stroke: useful predictor or goal of therapy? Age Ageing. 1989;18:183–191.
Wade DT, Hewer RL. Functional abilities after stroke: measurement, natural history and prognosis. J Neurol Neurosurg Psychiatry. 1987;50:177–182.
Wade DT, Wood VA, Hewer RL. Recovery after stroke: the first 3 months. J Neurol Neurosurg Psychiatry. 1985;48:7–13.
Anderson CS, Jamrozik KD, Broadhurst RJ, Stewart-Wynne EG. Predicting survival for 1 year among different subtypes of stroke. Stroke. 1994;25:1935–1944.
Stewart J, Dundas R, Howard RS, Rudd AG, Wolfe CDA. Ethnic differences in incidence of stroke: prospective study with stroke register. BMJ. 1999;318:967–971.
WHO Task Force on Stroke and Other Cerebrovascular Disorders. Stroke–1989: recommendations on stroke prevention, diagnosis, and therapy: report of the WHO Task Force on Stroke and Other Cerebrovascular Disorders. Stroke. 1989;20:1407–1431.
Rankin J. Cerebral vascular accidents in patients over 65, II: prognosis. Scot Med J. 1957;2:200–215.
Wood-Dauphinee SL, Williams JI, Shapiro SH. Examining outcome measures in a clinical study of stroke. Stroke. 1990;21:731–739.
Wolfe CDA, Taub NA, Woodrow J, Burney PGJ. Assessment of scales of disability and handicap for stroke patients. Stroke. 1991;22:1242–1244.
Brittain KR, Peet SM, Castleden CM. Stroke and incontinence. Stroke. 1998;29:524–528.
Borrie MJ, Campbell AJ, Caradoc-Davies TH, Spears GF. Urinary incontinence after stroke: a prospective study. Age Ageing. 1986;15:177–181.
Jawad SH, Ward AB, Jones P. Study of the relationship between pre-morbid urinary incontinence and stroke functional outcome. Clin Rehabil. 1999;13:447–452.