CHOLINESTERASE INHIBITOR TREATMENT AND URINARY INCONTINENCE IN ALZHEIMER'S DISEASE

Abstract

To the Editor: Dementia and urinary incontinence are common, often coexisting, problems in older people. Muscarinic receptors are important for detrusor contraction,1 and anticholinergic drugs are used to treat detrusor instability. Conversely, cholinesterase inhibitors, used to treat Alzheimer's disease (AD), have been reported to cause urinary incontinence.2 A retrospective population-based study of patients with dementia found that those prescribed cholinesterase inhibitors were significantly more likely to be subsequently prescribed an anticholinergic drug for urinary incontinence.3 Of the cholinesterase inhibitors currently prescribed for AD, only donepezil lists urinary incontinence as a potential adverse effect; rivastigmine and galantamine just mention the possibility of urinary obstruction.4 The potential to worsen urinary continence is an important consideration when starting cholinesterase treatment in AD. Urinary symptom data were collected from 216 (55 male, 161 female) consecutive patients attending a memory treatment center (mean age 76.7, range 49–94).5 Patients were eligible if they had a diagnosis of probable AD, were treated with a cholinesterase inhibitor, and had continence and cognitive assessments at baseline and after 26 weeks of treatment. Mean Mini-Mental State Examination (MMSE) score was 19.3 (n=213) at baseline and 18.8 (n=185) at 26 weeks. One hundred forty-two (65.7%) were treated with donepezil and 74 (34.3%) with rivastigmine. Urinary continence was scored as an item from the Physical Self-Maintenance Scale6 as fully continent, occasionally incontinent (no more than once a week), frequently incontinent, or always incontinent. Behavioral and psychological symptoms of dementia (BPSDs) were scored using the Neuropsychiatric Inventory (NPI).7 Cognitive testing comprised the MMSE, the Hopkins Verbal Learning Test,8 Verbal Fluency Categorical and Lexical forms,9 and for a proportion of patients, Paired Associate Learning and Delayed Matching to Sample subtests from the Cambridge Neuropsychological Test Automated Battery visual and working memory battery.10 Premorbid intelligence was estimated using the National Adult Reading Test (see 6). At baseline, 197 patients were fully continent, 12 had infrequent incontinence, one had frequent incontinence, and six were always incontinent. At 26 weeks, 184 were fully continent (a decrease of 6.6%), 18 had infrequent incontinence, three had frequent incontinence, and 11 were always incontinent. The change toward greater incontinence was significant (P=.03). Worsening continence was not associated with sex (chi-square=0.04, P=.85) or drug prescribed (chi-square=2.29, P=.13) on univariate analysis. A logistic regression model comprising age, sex, and drug prescribed with stepwise conditional forward entry at P<.05 found that rivastigmine significantly increased the risk of worse continence (odds ratio (OR)=2.91, 95% confidence interval (CI)=1.06–7.99). Adding baseline MMSE and NPI scores to the model (n=207) identified NPI as the only significant predictor (OR=1.05, 95% CI=1.01–1.09 per point higher). Including the other cognitive test scores in addition failed to improve the model significantly. To examine whether worse continence was associated with change in cognition or behavioral symptoms, the differences in MMSE and NPI scores between baseline and Week 26 were calculated. Entering these into the logistic regression model together with age, sex, and drug prescribed (n=179) identified poorer MMSE (OR=1.17, 95% CI=1.03–1.34 per point decline) and poorer NPI (OR=1.09, 95% CI=1.04–1.18 per point worse) as the only significant dependent variables. Cholinesterase inhibitor treatment was associated with significantly worse urinary continence in this sample. There was no significant difference between rivastigmine and donepezil once baseline characteristics were adjusted for. Patients with worse BPSDs were more likely to experience deterioration in continence; age, sex, and cognitive status were not significant predictors. When the course of cognitive and behavioral change was examined, significant associations between worse continence and cognitive and behavioral decline were found. This implies that improved cognition and behavior in patients who respond to cholinesterase inhibitors offsets the potential for worse continence. The incontinence rates for patients taking donepezil in this sample were the same as those found in a Japanese sample.2 They are higher than the proxy rate of anticholinergic prescriptions of 4% found in the Ontario dementia cohort.3 The proxy measure probably underestimates incontinence rates, because not all patients with incontinence receive anticholinergic treatments. Ideally, changes in continence should be assessed in future clinical trials of cholinesterase inhibitors. Until such trial data become available, clinicians should advise patients with AD considering cholinesterase inhibitor treatment that there is an approximately 7% risk of precipitating urinary incontinence and that any current incontinence may be significantly worsened. Risk is highest in those with more-severe BPSD, although should the patient respond well to treatment cognitively or behaviorally, the risk is substantially lower. Financial Disclosure: Grant support: Health Foundation, Leading Practice Through Research award, October 2003 through September 2005, to assess physical health status in older adults with learning disability. Author Contributions: JMS proposed the hypotheses, analyzed the data, and wrote the letter. Sponsor's Role: NHS Lothian Health Board approved and resourced the collection of all data as part of the evaluation of local dementia services.