Huang, Alison J. MD, MPhil1,2; Brown, Jeanette S. MD1,2; Thom, David H. MD, PhD2; Fink, Howard A. MD3; Yaffe, Kristine MD1,2; for the Study of Osteoporotic Fractures Research Group
Urinary incontinence is a common problem in older women, with up to 50% of women aged 60 years and older having incontinence at least once per week.1,2 Cognitive decline is also widespread in older women, occurring in up to 10% of persons aged 60 years and older, with women having a slightly higher risk than men.3,4 Among debilitated, institutionalized elderly women, incontinence is often seen in the setting of advanced dementia from conditions such as Alzheimer's disease and stroke.5–7 Among older women in the community, however, the association between milder, preclinical cognitive decline and incontinence is unclear.
Only a few studies have suggested that older community-dwelling adults with decreased cognitive function are more likely to have incontinence than those with preserved function.8,9 However, these studies did not assess important confounders and, in particular, did not assess physical function decline, which is correlated with cognitive decline in the elderly.10 Functional imaging studies have suggested that abnormalities in the frontal cerebral cortex may be associated with bladder control problems in older adults.11–14 However, the frontal cortex areas involved may be implicated not only in cognitive but also motor function.
At this time, it is unclear whether cognitive decline is independently associated with urinary incontinence in older community-dwelling women, compared with being a marker for other neurological or physical processes that contribute to this problem. We examined the association between cognitive decline, physical function decline, and incontinence in a large cohort of older community-dwelling women. We also estimated whether women with recent cognitive or physical decline are more likely to have incontinence that is disruptive to their daily activities.
MATERIALS AND METHODS
This analysis used data from the Study of Osteoporotic Fractures, a prospective cohort study of risk factors for fractures in older community-dwelling women. From September 1986 to October 1988, 9,704 women aged 65 years and older were recruited from population-based listings from four areas of the United States: Baltimore, Maryland; Minneapolis, Minnesota; the Monogahela Valley, Pennsylvania; and Portland, Oregon. Women were reassessed approximately every 2 years, with the first three follow-up visits occurring in 1989–1990 (visit 2), 1991–1992 (visit 3), and 1992–1994 (visit 4). The study was approved by the institutional review boards of all participating institutions.
Our analysis was restricted to participants who contributed data on cognitive function, physical function, and urinary incontinence through visit 4. Between the baseline visit and visit 4, 933 of the original Study of Osteoporotic Fractures enrollees died, and 142 terminated enrollment. Of those still enrolled at visit 4, 285 women did not return questionnaire data on incontinence, and an additional 1,882 women did not complete cognitive or physical function testing at baseline or visit 4, although they completed questionnaires. We also excluded 101 women who had abnormal cognitive function at baseline, defined as a modified Mini-Mental State Examination score of less than 20 out of 26 at the screening visit. Women with abnormal physical function at baseline, defined as an inability to walk without assistance, were already excluded from participating in Study of Osteoporotic Fractures. A total of 6,361 participants met our inclusion criteria. Compared with women who met inclusion criteria, excluded women were slightly older (mean age=71.8 compared with 71.5 years at baseline, P=.04) and more likely to have undergone hysterectomy (41% compared with 39%, P=.05), but did not differ significantly in education, overall health status, smoking, alcohol use, parity, body mass index, stroke, diabetes, or Parkinson disease (P<.1 for all). This analysis was approved by the institutional review board of the University of California San Francisco.
Cognitive function was measured in the Study of Osteoporotic Fractures using 3 tests: a modified version of the Mini-Mental State Examination (modified Mini-Mental State Examination), the Trails B test, and the Digit Symbol Substitution Test. These tests were selected 20 years ago as validated, easy-to-administer tests that would not create undue participant or site burden in this cohort.15 The modified Mini-Mental State Examination is short version of the standard Mini-Mental State Examination and is calculated out of 26 points, with higher scores corresponding to higher overall cognitive function. Trails B is a timed, written test requiring subjects to connect an alternating sequence of letters and numbers; shorter test times correspond to better executive function, attention, and visual scanning. The Digit Symbol Substitution Test is another timed, written test requiring subjects to translate numbers into symbols using a key; higher scores indicate greater psychomotor speed, attention, and perceptual organization. The modified Mini-Mental State Examination was administered at study visits 1 and 4, whereas the Trails B and Digit Symbol Substitution Test were administered at visits 2 and 4. To assess for recent cognitive decline, we examined change in cognitive function from visit 1 or 2 to visit 4. Women were considered to have recent, significant cognitive decline if their performance on a test declined more than one standard deviation (SD) beyond the mean change for the entire cohort. This approach is consistent with previous cognitive research supporting the use of population-based cutoff scores, in which clinically significant changes have been set at around 1.0 to 1.5 standard deviations beyond the mean change for a population.16,17
Physical function was assessed at visits 1 and 4 by examination of walking and chair stand speed, selected for the Study of Osteoporotic Fractures because they measure maneuvers important for daily living.18,19 Walking speed was measured by asking participants to walk at usual speed over a 6-meter course; the average of two trials was recorded to maximize accuracy (reliability, r=0.84). Chair stand speed was assessed by measuring the number of seconds needed to rise from a seated to a standing position five times in a row (reliability, r=0.70). We examined change in walking or chair stand speed from visit 1 to visit 4. Women were considered to have recent, significant physical function decline if their performance on either test worsened by more than 1 SD beyond the mean change.
Clinical frequency and functional disruptiveness of urinary incontinence were assessed in all Study participants by self-administered written questionnaire at visit 4. Women were asked, “During the last 12 months, have you ever leaked urine or lost control of your urine?” Women who reported any urine leakage in the past 12 months were asked, “How often does this leakage of urine usually occur—daily, one or more times per week but not every day, one or more times per month but not every week, less than once a month, or don't know?” to establish clinical frequency of symptoms. To assess functional disruptiveness of incontinence, participants were asked, “Does this leakage interfere with your activities?” Our primary outcomes were 1) “clinically frequent” incontinence, defined as urine leakage occurring at least once per week and 2) “functionally disruptive” incontinence, defined as leakage that interfered with activities.
Information about participants' medical and obstetric history and health-related habits was gathered by self-report questionnaire. Most data were obtained from visit 4, at the same time that incontinence was assessed; however, history of congestive heart failure and hysterectomy were available only from visit 2. Questionnaires were checked by interviewers at each study site for missing data; where data were incomplete, interviewers prompted participants to provide missing information. Overall health status was assessed by asking participants, “Compared to people your own age, how would you rate your overall health?” with response categories given as “excellent,” “very good,” “good,” “fair,” and “poor.”20 Depression was evaluated using the short form of the Geriatric Depression Scale, a self-administered screening instrument validated in elderly populations, in which scores of 6 or more are consistent with depression.21 Height and weight were both measured by examination at visit 4, and body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared.
Multivariable logistic regression was used to examine the associations between 1) significant cognitive decline and 2) significant physical function decline with the outcome of urinary incontinence. Because incontinence symptoms were assessed only at visit 4, we assessed the relationship of recent cognitive and physical decline to prevalent incontinence at visit 4. To develop our multivariable models, we first searched for univariable associations between potential confounders and our predictor and outcome variables. Covariates associated with both incontinence and either cognitive or physical decline at P<.2 in univariable analysis were entered into the appropriate multivariable model. Models were also adjusted for baseline cognitive or physical function. Hosmer-Lemeshow goodness-of-fit tests were used to confirm the appropriateness of each of our multivariable-fitted logistic regression models. No major violations of the assumptions of logistic regression were detected (P>.2 for all tests).
To avoid oversimplification associated with dichotomization of our predictor variables, we also performed an exploratory analysis in which we divided our cohort into quintiles defined by the magnitude of their recent change in cognitive and physical function (measured by modified Mini-Mental State Examination and by walking speed). We then assessed for a linear trend in the adjusted odds of weekly incontinence across the different quintiles of cognitive decline and physical function decline, using a test that evaluates orthogonal contrast in the log odds of the outcome for each of the categories.22 Finally, we used polytomous regression to estimate the relative risks of cognitive decline and physical function decline among women with a broader spectrum of incontinence frequency.23 Women reporting daily, weekly, or monthly incontinence were compared with women reporting less than monthly incontinence as the reference category. We then assessed for a linear trend in the relative risk of daily, weekly, and monthly incontinence in participants with recent cognitive decline and physical function decline.22 All analysis were performed using STATA 9.0 (StataCorp LP, College Station, TX).
Demographic and clinical characteristics of the 6,361 participants are summarized in Table 1. The mean age of women at visit 4 was 76.7±4.7 years. More than 80% of participants reported either excellent or very good overall health. The prevalence of known neurologic disease in this cohort was low, with 5% of women reporting any previous stroke, and less than 1% of women reporting Parkinson disease.
Just more than one half of women at visit 4 reported some incontinence in the past 12 months, and approximately 31% of women reported at least weekly incontinence. Only 5% of women (n=289) reported “disruptive” incontinence that interfered with their activities. There was a strong association between the frequency and disruptiveness of incontinence, in that 21% (n=178) of women with daily symptoms indicated that their incontinence interfered with activities, compared with only 4% (n=26) of those with monthly symptoms (P<.001). Nevertheless, approximately 12% (n=34) of women with “disruptive” incontinence had only monthly or less frequent symptoms, suggesting that frequency alone did not determine whether incontinence was disruptive.
On average, cognitive and physical function for the overall cohort worsened between baseline and visit 4. The mean (± standard deviation) change in modified Mini-Mental State Examination, Trails B, and Digit Symbol Substitution Test scores was –0.3±2.3, 18.6±50.1 seconds, and –1.4±16.2, respectively. The mean change in walking speed was –0.07±0.32 m/s, and the mean change in chair stand time was –0.4±6.5 seconds. For all tests, change in performance for the overall cohort was roughly normally distributed.
No significant association between recent cognitive decline and weekly incontinence was detected in either age-adjusted or multivariable-adjusted models (Table 2). This was true regardless of the specific test used to assess cognitive function. In contrast, recent physical function decline, as measured by decline in either walking speed or chair stand speed greater than 1 SD beyond the mean, was significantly associated with weekly incontinence, even after adjusting for age, health status, diabetes, geriatric depression score, BMI, stroke, alcohol use, and baseline physical function. On average, for example, women whose walking speed had declined more than 1 SD beyond the group mean had a 31% greater odds of having weekly incontinence (95% CI 9–56%) compared with women whose walking speed had decline less than 1 SD beyond the mean.
When we divided women into quintiles defined by the magnitude of recent cognitive decline, we did not find a significant linear trend in the odds of weekly incontinence associated with increasing quintile of either modified Mini-Mental State Examination or Trails B (P for trend>.1 for both), although there was a trend associated with increasing quintile of decline in Digit Symbol Substitution Test (P for trend=.006). In contrast, there was a strong trend in the odds of weekly incontinence associated with increasing quintile of physical decline, regardless of the specific test used (P for trend<.001 for both walking speed and chair stand decline).
Using polytomous regression to compare women with more subtle differences in the frequency of incontinence, we detected a strong linear trend in the association between recent physical function decline and frequency of incontinence (Fig. 1). Among women with recent decline in walking speed, for example, the risk of having monthly, weekly, and daily urinary incontinence was increased by an average of 3%, 22%, and 42%, respectively, after adjusting for age, diabetes, geriatric depression score, BMI, stroke, alcohol use, and baseline physical function (P=.002 for linear trend). No such linear trend was detected for the association between recent cognitive decline and increasing incontinence frequency (P>.1 for trend for all adjusted cognitive models).
Cognitive decline, as measured by either the modified Mini-Mental State Examination or Digit Symbol Substitution Test, was significantly associated with “disruptive” incontinence, after adjusting for age, health status, diabetes, geriatric depression score, stroke, alcohol use, Parkinson disease, frequency of incontinence, and baseline cognitive function (Table 3). For example, women with recent cognitive decline by modified Mini-Mental State Examination and Digit Symbol Substitution Test had an average 55% and 53% greater odds of disruptive incontinence, respectively, compared with women without cognitive decline. Furthermore, the association between decline in modified Mini-Mental State Examination and Digit Symbol Substitution Test and “disruptive” incontinence persisted even after controlling additionally for differences in physical function decline. That is, even after additional adjustment for decline in walking speed, women with recent decline in modified Mini-Mental State Examination score still had a 54% greater odds of having “disruptive” incontinence (95% CI 10–116%, P=.012). Women with decline in Digit Symbol Substitution Test score still had a 52% greater odds of having “disruptive” incontinence (95% CI 1–130%, P=.05). Although decline in Trails B score was associated with “disruptive” incontinence in age-adjusted analysis, this association was no longer significant after adjustment for other comorbidities, including physical function decline (OR 1.23, 95% CI 0.84–1.80, P=.294).
Decline in walking speed and decline in chair stand speed were significantly associated with “disruptive” incontinence in age-adjusted models (Table 3). However, these associations were attenuated and no longer significant after adjustment for other comorbidities, including cognitive decline (OR 1.42, 95% CI 0.97–2.08, P=.075, for women with walking speed decline; OR 1.20, 95% CI 0.85–1.68, P=.297, for women with decline in chair stand speed).
We estimated the proportion of women whose “clinically frequent” or “functionally disruptive” incontinence was attributable to physical function decline (by walking speed) or cognitive decline (by modified Mini-Mental State Examination), respectively, using the equation: attributable risk proportion=prevalence × (RR – 1)/[prevalence × (RR – 1) + 1]. Physical function decline accounted for an estimated 9% of clinically frequent incontinence, whereas cognitive decline accounted for 3% of disruptive incontinence in this population.
This study provides important insight into the relationships among cognitive decline, physical function decline, and self-reported urinary incontinence in older women. In this cohort of generally healthy, community-dwelling elderly women, women with recent cognitive decline were no more likely to have weekly incontinence than those without cognitive decline, regardless of the test used to assess cognitive function. On the other hand, women with recent physical function decline, as measured by decline in both walking speed and chair stand speed, were substantially more likely to report weekly incontinence, even after adjusting for age, health status, and comorbid conditions. These findings suggest that physical function decline may be a stronger risk factor for having clinically frequent incontinence than cognitive decline in older community-dwelling women.
At the same time, our findings also indicate that cognitive decline may be an important risk factor for having functionally “disruptive” incontinence. That is, even after adjusting for differences in the frequency of incontinence, women with recent decline in modified Mini-Mental State Examination or Digit Symbol Substitution Test score were significantly more likely to report that their urine leakage interfered with their activities than those without decline. Furthermore, the association between cognitive decline and “disruptive” incontinence persisted even after controlling for differences in physical function decline.
At first glance, it seems puzzling that cognitive decline should be a marker for having functionally “disruptive” incontinence, but not for weekly incontinence. One possible explanation is that “disruptive” incontinence represents a subset of especially clinically severe incontinence, one that cannot be adequately described using standard frequency categories such as “daily,” “weekly,” or “monthly” urine leakage. Another possibility, however, is that women with recent cognitive decline may have more difficulty coping with or adjusting to incontinence than women without cognitive decline, even when their symptoms are equally frequent or severe. This suggests that, at least in the nondemented elderly, cognitive decline is not so much a risk factor for the development of incontinence as a comorbid factor that magnifies the impact of incontinence on functioning and quality of life. Further research is needed to determine whether different cognitive domains may be more or less implicated in both clinical frequency and functional disruptiveness of incontinence in older women.
Although several previous studies have pointed to an association between cognitive impairment and incontinence in older community-dwelling adults, none examined either incontinence or cognitive function in much detail, and adjustment for confounders was limited. One study of mostly community-dwelling adults aged 85 years and older in Sweden, for example, found that incontinence was associated with self-reported diagnosis of dementia.8 The relationship between incontinence and milder forms of cognitive impairment was not examined, however, and other comorbidities were not examined. Another large cross-sectional study of community-dwelling persons aged older than 75 years in Great Britain found that cognitive impairment as defined by a Mini-Mental State Examination score less than 25 was associated with a 30% increased odds of having incontinence.9 The authors did not characterize the duration and frequency of incontinence in this population and did not take into account comorbidities other than history of falls and vision or hearing problems. Other studies have not considered community-dwelling individuals separately from institutionalized persons,14,24 despite marked differences in the spectrum of mental and physical function of these populations.
Our findings have important implications for the evaluation and management of urinary incontinence in older women living in the community. Although the prevalence of recent cognitive decline may be similar in women with and without incontinence, those women with recent cognitive decline may be more likely to find their symptoms to be disruptive. Clinicians concerned about the effect of incontinence on women's daily lives may want to consider formally screening for cognitive impairment in those who have especially disruptive incontinence symptoms. This issue may be particularly relevant for patients taking anticholinergic medications to treat urge- or mixed-type incontinence, because these agents have the potential to worsen cognitive function in the elderly by promoting delirium.25–27
This research has several important limitations. Because incontinence was assessed at visit 4 only, we could not assess for differences in the incidence of incontinence. As a result, we could not assess whether cognitive or physical function decline played a causal role in the development of urinary incontinence over time. Second, the questionnaire used at visit 4 did not distinguish between stress, urge, or mixed incontinence symptoms, even although there are important differences in these clinical types of incontinence.28–30 It is possible that decline in physical function might cause more problems for women with incontinence related to a sudden urge to urinate, compared with incontinence that is reliably brought on by maneuvers associated with changes in abdominal pressure. Third, the disruptiveness of women's incontinence symptoms was assessed using a single-item measure, and further research using more detailed instruments may help to provide further insight into the way that cognitive and physical function decline affect older women's abilities to cope with incontinence.
Finally, it is possible that self-report of incontinence may not be as reliable in older persons with cognitive decline, compared with persons with preserved cognitive function. If cognitive decline were to cause women to under-report their incontinence symptoms, this might result in an underestimate of the association between cognitive decline and incontinence in our study. However, in this cohort of community-dwelling women, in which the spectrum of cognitive decline was relatively mild, we would not expect women's ability to distinguish between broad categories of incontinence frequency, or to report incontinence that was disruptive to their activities, to be significantly compromised.
In conclusion, both cognitive and physical function decline are associated with urinary incontinence in older community dwelling women. Although recent physical function decline is more strongly correlated with having more frequent incontinence, cognitive decline may be more strongly associated with functionally “disruptive” incontinence that interferes with activities. Clinicians may want to consider formally assessing cognitive function in older women presenting with incontinence.
1. Grodstein F, Fretts R, Lifford K, Resnick N, Curhan G. Association of age, race, and obstetric history with urinary symptoms among women in the Nurses' Health Study. Am J Obstet Gynecol 2003;189 (2):428–34.
2. Thom D. Variation in estimates of urinary incontinence prevalence in the community: effects of differences in definition, population characteristics, and study type. J Am Geriatr Soc 1998;46:473–80.
3. Manly JJ, Bell-McGinty S, Tang MX, Schupf N, Stern Y, Mayeux R. Implementing diagnostic criteria and estimating frequency of mild cognitive impairment in an urban community. Arch Neurol 2005;62:1739–46.
4. Evans DA, Funkenstein HH, Albert MS, Scherr PA, Cook NR, Chown MJ, et al. Prevalence of Alzheimer's disease in a community population of older persons. Higher than previously reported. JAMA 1989;262:2551–6.
5. Ouslander JG, Zarit SH, Orr NK, Muira SA. Incontinence among elderly community-dwelling dementia patients. Characteristics, management, and impact on caregivers. J Am Geriatr Soc 1990;38:440–5.
6. Ouslander JG, Uman GC, Urman HN, Rubenstein LZ. Incontinence among nursing home patients: clinical and functional correlates. J Am Geriatr Soc 1987;35:324–30.
7. Nelson RL, Furner SE. Risk factors for the development of fecal and urinary incontinence in Wisconsin nursing home residents. Maturitas 2005;52:26–31.
8. Hellstrom L, Ekelund P, Milsom I, Skoog I. The influence of dementia on the prevalence of urinary and faecal incontinence in 85-year-old men and women. Arch Gerontol Geriatr 1994;19:11–20.
9. Rait G, Fletcher A, Smeeth L, Brayne C, Stirling S, Nunes M, et al. Prevalence of cognitive impairment: results from the MRC trial of assessment and management of older people in the community. Age Ageing 2005;34 (3):242–8.
10. Rosano C, Simonsick EM, Harris TB, Kritchevsky SB, Brach J, Visser M, et al. Association between physical and cognitive function in healthy elderly: the health, aging and body composition study. Neuroepidemiology 2005;24:8–14.
11. Blok BF, Willemsen AT, Holstege G. A PET study on brain control of micturition in humans. Brain 1997;120:111–21.
12. Griffiths D. Clinical studies of cerebral and urinary tract function in elderly people with urinary incontinence. Behav Brain Res 1998;92:151–5.
13. Griffiths D, Derbyshire S, Stenger A, Resnick N. Brain control of normal and overactive bladder. J Urol 2005;174:1862–7.
14. Griffiths DJ, McCracken PN, Harrison GM, Gormley EA, Moore K, Hooper R, et al. Cerebral aetiology of urinary urge incontinence in elderly people. Age Ageing 1994;23:246–50.
15. Yaffe K, Grady D, Pressman A, Cummings S. Serum estrogen levels, cognitive performance, and risk of cognitive decline in older community women. J Am Geriatr Soc 1998;46:816–21.
16. Ganguli M, Belle S, Ratcliff G, Seaberg E, Huff FJ, von der Porten K, et al. Sensitivity and specificity for dementia of population-based criteria for cognitive impairment: the MoVIES project. J Gerontol 1993;48:M152–61.
17. Gregg EW, Yaffe K, Cauley JA, Rolka DB, Blackwell TL, Narayan KM, et al. Is diabetes associated with cognitive impairment and cognitive decline among older women? Study of Osteoporotic Fractures Research Group. Arch Intern Med 2000;160:174–80.
18. Nelson HD, Nevitt MC, Scott JC, Stone KL, Cummings SR. Smoking, alcohol, and neuromuscular and physical function of older women. Study of Osteoporotic Fractures Research Group. JAMA 1994;272:1825–31.
19. Sands ML, Schwartz AV, Brown BW, Nevitt MC, Seeley DG, Kelsey JL. Relationship of neurological function and age in older women. The study of osteoporotic fractures. Neuroepidemiology 1998;17:318–29.
20. Lundberg O, Manderbacka K. Assessing reliability of a measure of self-rated health. Scand J Soc Med 1996;24:218–24.
21. Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982;17:37–49.
22. Vittinghoff E, Glidden D, Shiboski S, McCulloch C. Regression methods in biostatistics: linear, logistic, survival, and repeated measures models. New York (NY): Springer;2005. p. 82–83.
23. Hosmer DW, Lemeshow S. Special topics. In: Applied logistic regression. New York (NY): Wiley; 1989. p. 260–351.
24. Coppola L, Caserta F, Grassia A, Mastrolorenzo L, Altrui L, Tondi G, et al. Urinary incontinence in the elderly: relation to cognitive and motor function. Arch Gerontol Geriatr 2002;35:27–34.
25. Scheife R, Takeda M. Central nervous system safety of anticholinergic drugs for the treatment of overactive bladder in the elderly. Clin Ther 2005;27:144–53.
26. Todorova A, Vonderheid-Guth B, Dimpfel W. Effects of tolterodine, trospium chloride, and oxybutynin on the central nervous system. J Clin Pharmacol 2001;41:636–44.
27. Katz IR, Sands LP, Bilker W, DiFilippo S, Boyce A, D'Angelo K. Identification of medications that cause cognitive impairment in older people: the case of oxybutynin chloride. J Am Geriatr Soc 1998;46:8–13.
28. Klingele CJ, Carley ME, Hill RF. Patient characteristics that are associated with urodynamically diagnosed detrusor instability and genuine stress incontinence. Am J Obstet Gynecol 2002;186:866–8.
29. Thom DH, van den Eeden SK, Ragins AI, Wassel-Fyr C, Vittinghof E, Subak LL, et al. Differences in prevalence of urinary incontinence by race/ethnicity. J Urol 2006;175:259–64.
30. Luber KM, Boero S, Choe JY. The demographics of pelvic floor disorders: current observations and future projections. Am J Obstet Gynecol 2001;184:1496–501.
© 2007 by The American College of Obstetricians and Gynecologists.