This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

New-Onset Incontinence and Markers of Frailty

Data From the Hispanic Established Populations for Epidemiologic Studies of the Elderly

^a Departments of Family and Community Medicine, University of Texas Health Science Center at San Antonio
^b Internal Medicine, University of Texas Health Science Center at San Antonio
^c Department of Preventive Medicine and Community Health, University of Texas Medical Branch at Galveston

Toni P. Miles, Department of Family and Community Medicine, 7703 Floyd Curl Drive, Mailbox 7795, San Antonio, TX 78229-3900 E-mail: milest{at}uthscsa.edu.

Decision Editor: John E. Morley, MB, BCh

	Abstract

Top Abstract Methods Results Discussion References

 
Background. Urinary incontinence in older adults has many distinct etiologies and is associated with lower self-reported health. However, it is unclear whether the new onset of urinary incontinence marks newly emergent frailty.

Methods. Using a longitudinal population-based survey of older Mexican Americans (N = 2660) across five southwestern states, this study compared the strength of association between markers of physical frailty such as activities of daily living (ADLs), instrumental activities of daily living (IADLs), and performance (timed walk, timed chair rise, and tandem balance) with baseline incontinence (prevalent disease) and new-onset incontinence (incident disease).

Results. We found that 14.1% of the participants (n = 329) were incontinent at baseline (prevalent cases) and 11.6% (n = 208) were newly incontinent 2 years later (incident cases). Controlling for other covariates, prevalent incontinence was only associated with a 60% increased risk of having difficulty walking 8 ft. Incident incontinence was associated with a twofold increased risk of impairment in ADLs and IADLs, and poor performance on all three physical measures.

Conclusions. Incident incontinence is associated with an increased risk of more global functional impairment. Thus, incident disease may be an important early marker for signaling the onset of frailty among persons who become incontinent after the age of 65 years.

INCONTINENCE is initiated by a number of events, with distinctly different causal mechanisms for its emergence, such as menopause ⁽¹⁾⁽²⁾, surgical procedures ^(3)(4)(5), and complications from a chronic condition ⁽⁶⁾ or in association with declines in cognitive function ⁽⁷⁾. With a few exceptions, most of these events are commonly found among persons aged 65 years and older, and their probability of occurrence increases with increasing age. In addition, incontinence is often associated with frailty ⁽⁸⁾⁽⁹⁾. There are many risk factors common to the development of both incontinence and frailty. These common factors include multiple medical conditions, depression, and poor self-assessed health ^{(6)(8)(10)(11)}. In this analysis, we present evidence suggesting that older persons who are newly incontinent may also be in the earliest stages of the development of frailty.

Among community-dwelling older adults, incontinence prevalence estimates can range from 2% to 55% and are largely dependent on the structure of survey items ⁽¹²⁾. Questions about "difficulty using the toilet" tend to yield lower prevalence rates than items that directly query the respondent concerning frequency of "difficulty holding urine" and intensity of "wet" clothing ^(10)(12). Surveys that use an additional introductory text and a follow-up probe obtain the greatest prevalence rates ⁽¹³⁾. Attempts have been made to estimate the prevalence of incontinence subtypes (urge, stress, overflow, and mixed) with survey questions, but these appear to be poorly correlated with urodynamic testing ⁽¹⁴⁾.

Frailty is a global concept that embodies elements of self-reported health, self-reported physical function measures, and physical performance measures ^(15)(16). There is considerable evidence linking frailty and incontinence. One global measure of frailty that has been associated with incontinence is the self-reported health item: "How would you rate your health?" ⁽¹⁷⁾. Activities of daily living (ADLs) and instrumental activities of daily living (IADLs) are self-reported physical measures that also have an established association with prevalent incontinence ⁽¹⁸⁾. When ADL and IADL scales are combined, the psychometric properties of the scale also suggest that there is a common latent trait captured by both of these scales, possibly frailty ⁽¹⁹⁾. In this single latent factor structure, all items have substantial loadings (.71). The incontinence item on the ADL scale also loads on this single factor (loading = .845). This indirect evidence suggests that ADLs, IADLs, and incontinence all capture common elements of frailty. Although it would be expected, no studies to date have established an association between incontinence and the ability to perform an 8-ft walk, repeated chair stands, or standing-balance postures. There is, however, good evidence linking prevalent disability in all ADL items and two IADL items (difficulty with a 1/2mile walk and climbing stairs) with poor performance in physical performance tests ⁽¹⁵⁾. If difficulty with ADLs, IADLs, and physical performance tasks are all present when frailty is fully developed, then the presence of a single difficulty with any of these tasks can be viewed as a precursor of frailty. It is reasonable, then, to propose that incident incontinence may be linked to the appearance of these frailty precursors.

In the present analysis, the appearance of incontinence in formerly continent older adults was evaluated for its association with measures of functional status. To test this specific association, we removed individuals from the sample who were already reporting episodes of incontinence. These cases represented a mixture of persons whose symptoms occurred before or after age 65 years, and whose disease was due to a variety of causes that may have had less to do with functional decline over a 2-year period.

Testing the strength of this association required several elements. First, we identified new-onset cases of incontinence. Next, we designed a summary measure of frailty by combining measures of functional status. Finally, we tested the independence of incident incontinence in a model designed to predict frailty, adjusted for more usual covariates. A similar model was tested substituting prevalent incontinence for comparison.

	Methods

Top Abstract Methods Results Discussion References

 
Sample
The Hispanic Established Population for Epidemiologic Studies of the Elderly (HEPESE) was the primary focus of this analysis. The HEPESE is a longitudinal survey of a representative sample of 3051 Mexican Americans, aged 65 and older, residing in the southwestern United States. The sample was drawn using area probability sampling procedures to represent the Mexican American elderly population in the five southwestern states of Texas, New Mexico, Colorado, Arizona, and California. The HEPESE adds to the ethnic and regional diversity of existing Established Population for Epidemiologic Studies of the Elderly (EPESE) cohorts ⁽²⁰⁾ and was designed for population-based studies of elderly community-dwelling Mexican Americans. Its complex sampling design also enables researchers to estimate population-level prevalence. The other EPESE sites included whites residing in the Northeast (New Haven, East Boston), Midwest (Iowa), and the Southeast (North Carolina), and African Americans residing in North Carolina and New Haven. Specific questions included in the baseline HEPESE survey were selected to compare with these four other EPESE sites ⁽²⁰⁾.

Participants were interviewed in their homes. The length of time between the initial and follow-up survey was 2 years (fall 1993–spring 1994 to fall 1995–spring 1996). The response rate for eligible respondents was 86%. In this analysis, subjects were excluded at either time period if their Mini-Mental State Examination score was less than 17 (n = 152) or if there were missing or incomplete data (n = 198). There were 2167 persons who completed the follow-up survey. The remainder died (224), required proxy respondents (272), refused to participate (109), or were lost to follow-up (278).

Denominators.-- During the course of this analysis, several samples had to be specified. In Table 1 , the baseline sample had 2660 persons, including 329 prevalent cases (having some level of difficulty holding urine). There were 2025 persons who had data available at both time points. Among these, 208 were newly incontinent. If we do not take baseline continence into account, then 326 persons are counted as prevalent cases at follow-up.

The distribution of scores for ADLs and IADLs was markedly skewed (as most respondents reported little or no impairment). After trying other transformation procedures to improve the distribution, we created dichotomous ADL and IADL variables where 0 indicates no need for assistance and 1 indicates need for assistance on at least one item.

Physical performance.-- All performance measures were obtained during the home interview. Participants were asked to perform a timed 8-ft walk, a repetitive chair stand, and to hold several different balance postures of escalating difficulty (semitandem, tandem, and single-leg stance) using a protocol previously described ⁽¹⁶⁾. Data collected in the home consisted of time taken to complete the task (walk and chair stands) or time that posture was held (balance).

A physical performance score was created based on an assessment of the participant's ability to complete the task in the analysis—not the amount of time taken to complete it. To create this physical performance score, each test was reclassified into one of the five categories: unable, poor, moderate, good, or best. For the 8-ft walk test: poor = 10.0 seconds; moderate = 6.0 to 9.0 seconds; good = 4.0 to 5.0 seconds; and best = 3.0 seconds. For the five timed chair stands, poor = >16.5 seconds; moderate = 13.6 to 16.4 seconds; good = 11.0 to 13.5 seconds; and best = 10.9 seconds. These times corresponded to quartiles for the entire cohort. For the balance tests, poor = both side by side and semitandem held for 10 seconds, but tandem held for <2 seconds; moderate = tandem position held between 2 and 10 seconds; and best = tandem position held for 10 seconds.

Frailty.-- Our summary frailty measure was used as an outcome (Table 3 ). To create this variable, scores from ADLs, IADLs, walk, balance, and chair were summed. The summary score exhibited an approximation of normality at both time points (baseline skew = .77, kurtosis = -.19; follow-up skew = .80, kurtosis = -.53). This approach meant that assessment of frailty was conservative, with an emphasis on the most severely affected within this community-dwelling sample.

The covariates selected for inclusion in the models were the Center for Epidemiologic Studies–Depression Scale ⁽²⁵⁾, self-reported number of medical conditions (maximum, five), self-reported health, body mass index (BMI), gender, age, history of smoking (at least 100 cigarettes in a lifetime), and baseline frailty. These covariates were selected primarily to control for the shared association each has with both incontinence and frailty. To disentangle the relationship of incontinence and frailty, we must control for these other covariates. It has been observed among most cohorts of older adults that women have a greater prevalence of both incontinence and frailty. To test the independence of incident incontinence as a risk factor for frailty, we held gender as a constant in all adjusted models (Table 3 ). Both depression and chronic illness have been associated with incontinence and poor physical function. In addition, both chronic illness and smoking are associated with declining health. Medical conditions were ascertained in items asking respondents to indicate if they had ever been diagnosed by a doctor with one of the following: cardiovascular disease (heart attack or stroke), hypertension, cancer, diabetes, or arthritis. Scores were 1 for yes and 0 for no. For the number of medical conditions, a summary variable was created by summing the individual items. Measured height and weight were used to calculate BMI (kg/m²). In the analysis, BMI was divided into three categories: <24 kg/m², 24 to 27 kg/m², and >27 kg/m².

Statistical Analysis
All analyses were performed using SUDAAN (release 6.40, Research Triangle Institute, NC), a software program that accounts for design effects and multi-stage sampling procedure ⁽²⁶⁾. Statistical testing for differences of percentages and means between baseline and follow-up was done using the chi-square test for categorical variables and paired t tests. Bivariate and multivariate ordinary least squares and logistic regression were used to estimate coefficients and to estimate odds ratios.

	Results

Top Abstract Methods Results Discussion References

 
In the weighted sample of 436,533 persons, there were 60,565 cases of incontinence at baseline (prevalent cases). Another 32,271 persons became incontinent between the baseline and follow-up periods (incident cases). These numbers are consistent with the 2-year incidence rates from other studies of community-dwelling older persons ⁽²⁷⁾. Table 1 shows the distribution of continence components, frailty components (ADLs, IADLs, and physical performance variables), and covariates used in analyses with the unweighted sample. The two prevalent incontinence rates shown for baseline and follow-up do not capture the seriousness of incident disease.

To estimate a true incidence of incontinence, we removed the following persons from the follow-up sample: those who were incontinent at both time periods (n = 107, 6%) and those who were only incontinent at baseline (n = 124, 7%). Among those who were continent at baseline, 11.6% (208) reported difficulty 2 years later. This 11.6% is substantially larger than the apparent difference when baseline (14.1%) and follow-up (16.1%) prevalence rates are compared, indicating a greater population burden of new-onset disease.

The phrasing of survey items can influence the estimated prevalence of incontinence. At baseline, five items were available to assess continence. The highest prevalence rates were obtained with queries about "Difficulty stopping flow" (42.8%) and "Need for protection" (32.4%). The item designed to detect spastic bladder ("Feel the need") was endorsed by 26.8% of the sample. The item "Difficulty holding urine" (14.1%) identified a low proportion compared with these items. An additional incontinence measure available in the survey was the ADL item "Do you have difficulty using the toilet?" At baseline, 2.9% indicated difficulty with this one item. At follow-up, 5.0% indicated difficulty with the task—a substantial increase over baseline. Of the other self-reported incontinence items, only one, "Difficulty holding urine," was available at both baseline and follow-up. Although this appears to be the least specific, there was a significant increase in the proportion of incontinent subjects between baseline and follow-up (14.1% vs 16.1%).

Table 2 was designed to facilitate the comparisons of prevalent or incident cases with continent persons interviewed during the same survey period. Distinctly different patterns of associated characteristics emerge when this distinction is made. Persons classified as prevalent cases are shown in column 1, with the comparison continent sample shown in column 2. With these two groups, it can be seen that prevalent cases are older, with significantly higher rates of depression, poor health, and numbers of chronic diseases. Cases are also more likely to have smoked and have difficulty with ADLs, IADLs, and performance tasks.

If, on the other hand, incident cases (column 3) are compared with continent persons (column 4), the pattern of differences changes. Newly incident persons are significantly heavier and have greater ADL and IADL difficulty. There is a trend toward greater difficulty with performance tasks that does not reach statistical significance. Age, depression, self-reported health, and number of chronic diseases are no longer significant.

Two models were developed to test both prevalent and incident incontinence as independent predictors of frailty. To create the dependent variable for both models, all the markers of frailty (ADLs, IADLs, 8-ft walk, standing balance, and repeated chair stands) were combined as a single measure. An adjusted and an unadjusted version of each model were evaluated. Model 1 used prevalent incontinence as a predictor variable. In the unadjusted model, which also included baseline frailty, prevalent disease is significant (R²= .29, p < .05). After adjustment for significant covariates (BMI, sex, age, smoking, and number of chronic diseases), the model fit with the data improved slightly (R²= .33), and prevalent disease lost its significance. This suggests that the association between frailty and prevalent disease is explained by the covariates (i.e., that prevalent incontinence is not an independent predictor of frailty).

A second model using the same analysis, but substituting incident incontinence as an independent variable, was fit to the data. In the unadjusted model incident, incontinence was a significant predictor (R²= .28). Model fit also improved with adjustment (R²= .33) for significant covariates, and incident incontinence remained significant. Although the improvement in fit was small, this suggests that incident incontinence is independently associated with frailty.

Are there differences in the strength association between individual components of frailty and incident incontinence? Table 3 shows the results of five logistic models targeting this question. In this analysis, each component of frailty (ADLs, IADLs, 8-ft walk, standing balance, and repeated chair stands) were evaluated. As in the previous models, each model was adjusted for age, sex, chronic disease, BMI, and lifetime smoking. Prevalent incontinence was associated with an almost 60% increased risk of having difficulty in the 8-ft walk. Persons with incident incontinence had a twofold increased risk for having difficulty with ADLs and IADLs, as well as poor performance on the three physical performance tasks. This suggests there may be a common process linking difficulty in lower body mobility with the development of incontinence. The emergence of IADLs suggests that the common process may have its origin in neural control mechanisms since IADL functions—like walking and balance—are integrated tasks.

Summary
New-onset incontinence and mild difficulty with lower body mobility appear to represent an interim stage in the progression toward frailty. What is the etiology of this combined incontinence and mobility difficulty? We cannot say with these data. There are multiple etiologies ranging from central nervous system and peripheral nervous system pathologies, musculoskeletal abnormalities to mechanical problems (extreme obesity to deconditioning). Each etiology would have a specific and very distinct therapeutic response. These causes more commonly occur after the age of 65. So, for example, although the estrogen deficit associated with menopause is less likely to be a cause of new-onset incontinence in this group, a mild stroke with residual deficits would be a more likely cause.

	Discussion

Top Abstract Methods Results Discussion References

 
It is important to understand that this sample is a community-dwelling group of older Mexican Americans in the southwestern United States, representative of the type of patient most commonly seen by primary care physicians. In the analysis, determination of incident disease was derived from differences in self-report at baseline and follow-up 2 years later. This interval is typical of the time between physician visits for healthy older adults ⁽²⁸⁾. However, our analysis may underestimate the strength of the association between incident incontinence and frailty markers, because the question used in this survey has been shown to underestimate the prevalence of incontinence. We do not have specific details surrounding the onset of incontinence. If incident cases were due to surgical procedures, a change in cognitive skills, or the addition of a new medication, we were unable to detect these as exogenous causes of a change in continence. Our baseline prevalence of incontinence is consistent with rates observed in other community-dwelling samples that used a similarly worded question ⁽¹²⁾. In this sample of Mexican Americans, we have observed different prevalence rates based on the wording of incontinence items. The differential response rates to these items suggests that work needs to be done to identify a more culturally specific survey item that captures the extent of incontinence among older Mexican Americans. Reports comparing blacks and whites suggest that variation in the interpretation of a survey question is an important contributor to observed group differences in prevalence rates ⁽¹³⁾. However, we are not aware of similar data for older Mexican Americans. This analysis is focused solely on incident incontinence. Cases of incontinence remission have been reported in other cohorts ⁽²⁷⁾, as well as in ours. Elucidating a causal mechanism for the development of incontinence and its resolution is beyond the limited data available in our study.

The use of combined self-reports and physical performance measures to create a frailty measure is a process that may complicate our analysis in ways that cannot be completely addressed. Ideally, we wanted to estimate the strength of incident incontinence against the risk of new frailty at follow-up. It should be remembered that in our regression model, both incontinence (incident and prevalent) and frailty at follow-up were obtained in the same survey. The model is probably detecting an association between new incontinence and either an incremental increase in frailty or the emergence of new symptoms. To identify new-onset incontinence as a precursor of any impairment, we need continued follow-up of those persons who start with new-onset incontinence and no mobility difficulty.

	Acknowledgments

 
This research was supported by the Hispanic Established Population for Epidemiologic Studies of the Elderly Grant 5R01AG 10939-08 and Health Aging and Body Changes National Institute on Aging Contracts N01-AG-2102, 2103, and 2106.

Received August 17, 2000

Accepted August 22, 2000

	References

Top Abstract Methods Results Discussion References

 

1. Burgio KL, Matthews KA, Engel BT, 1991. Prevalence, incidence and correlates of urinary incontinence in healthy, middle-aged women. J Urol. 146:1255-1259. [Medline]
2. Brown JS, Grady D, Ouslander JG, Herzog AR, Varner E, Posner SF, 1999. Prevalence of urinary incontinence and associated risk factors in postmenopausal women. Obstet Gynecol 94:66-70. [Medline]
3. Bishoff JT, Motley G, Optenberg SA, et al. 1998. Incidence of fecal and urinary incontinence following radical perineal and retropubic prostatectomy in a national population. J Urol 160:454-458. [Medline]
4. Lepor H, Williford WO, Barry MJ, Haakenson C, Jones K, 1998. The impact of medical therapy on bother due to symptoms, quality of life and global outcome and factors predicting response. J Urol 160:1358-1367. [Medline]
5. Thom DH, Brown JS, 1998. Reproductive and hormonal risk factors for urinary incontinence in later life: a review of the clinical and epidemiologic literature. J Am Geriatr Soc 46:1411-1417. [Medline]
6. Black SA, Ray LA, Markides KS, 1999. The prevalence and health burden of self-reported diabetes in older Mexican Americans: findings from the Hispanic Established Populations for Epidemiologic Studies of the Elderly. Am J Public Health. 89:546-552. [Abstract/Free Full Text]
7. Brittain KR, Peet SM, Castleden CM, 1998. Stroke and incontinence. Stroke 29:524-528. [Abstract/Free Full Text]
8. Johnson TM, Kincade JE, Bernard SL, Busby-Whitehead J, Picciotto-Hertz I, DeFriese GH, 1998. The association of urinary incontinence with poor self-rated health. J Am Geriatr Soc 46:693-699. [Medline]
9. Dunn JE, Furner SE, Miles TP, 1993. Do falls predict institutionalization in older persons?. J Aging Health. 5:194[Abstract/Free Full Text]
10. Herzog AR, Fultz NH, 1990. Prevalence and incidence of urinary incontinence in community-dwelling populations. J Am Geriatr Soc 38:273-281. [Medline]
11. Ekelund P, Grimby A, Milson I, 1993. Urinary incontinence social and financial costs high. Br Med J 306:1344-1345.
12. Thom DH, 1998. Variation in estimates of urinary incontinence prevalence in the community: effects of differences in definition, population characteristics, and study type. J Am Geriatr Soc. 46:473[Medline]
13. Fultz NH, Herzog AR, 2000. Prevalence of urinary incontinence in middle-aged and older women: a survey-based methodological experiment. J Aging Health. 12:459-469. [Abstract/Free Full Text]
14. Kirschner-Hermanns R, Scherr PA, Branch LG, Wetle T, Resnick NM, 1998. Accuracy of survey questions for geriatric urinary incontinence. J Urol 159:1903-1908. [Medline]
15. Guralnik JM, Simonsick EM, Ferrucci L, et al. 1994. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol Med Sci 49:M85-M94.
16. Ostir GV, Markides KS, Black SA, Goodwin J, 1998. Lower body function is a predictor of subsequent disability among older Mexican Americans. J Gerontol Med Sci 53A:M491-M495. [Abstract]
17. Roberts RB, Jacobsen SJ, Rhodes T, et al. 1998. Urinary incontinence in a community-based cohort: prevalence and healthcare seeking. J Am Geriatr Soc 46:467[Medline]
18. Brandeis G, Bauman M, Hossain M, Morris J, Resnick N, 1997. Prevalence of potentially remediable urinary incontinence in frail older people: a study using the Minimum Data Set. J Am Geriatr Soc 45:179[Medline]
19. Spector WD, Fleishman JA, 1998. Combining activities of daily living with instrumental activities of daily living to measure functional disability. J Gerontol Soc Sci 53A: (suppl) S46-S57. [Abstract]
20. Cornoni-Huntley J, Blazer DG, Lafferty ME, Everett DF, Brock DB, Farmer ME, National Institutes of Health; 1990Publication 90–495.. Established Populations for Epidemiologic Studies of the Elderly, Vol. II, Resource Data Book National Institute on Aging, Bethesda, MD.
21. Katz S, Ford AB, Moskowitz RW, 1963. Studies of illness in the aged; the index of ADL: a standardized measure of biological and psychosocial function. JAMA 185:914
22. Nagi SZ, 1976. An epidemiology of disability among adults in the United States. Milbank Q 54:439-467.
23. Hayward MD, Crimmins EM, Miles TP, Yang Y. The Significance of Socioeconomic Status in Explaining the Race Gap in Chronic Health Conditions. In press.
24. Rozenzweig A, Prigerson H, Miller MD, Reynolds CF, 1997. Bereavement and late-life depression: grief and its complications in the elderly. Ann Rev Med. 48:421-428. [Medline]
25. Radloff LS, 1977. The CES-D scale: a self-report depression scale for research in the general population. J Appl Psychol Meas. 1:385
26. Shah BV, Barnwell BG, Bieler GS. SUDAAN ^TM User's Manual: Software for Analysis of Correlated Data, Release 6.40. Research Triangle Park, NC: Research Triangle Institute; 1995.
27. Herzog AR, Diokno AC, Brown MB, Fultz N, Goldstein N, 1994. Two-year incidence, remission, and change patterns of urinary incontinence in noninstitutionalized older adults. J Am Geriatr Soc 42:264-268. [Medline]
28. U.S. Senate Committee on Aging. Aging America. Washington, DC: U.S. Senate Committee on Aging; 1991.

This article has been cited by other articles:

				K. Bandeen-Roche, Q.-L. Xue, L. Ferrucci, J. Walston, J. M. Guralnik, P. Chaves, S. L. Zeger, and L. P. Fried Phenotype of Frailty: Characterization in the Women's Health and Aging Studies. J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2006; 61(3): 262 - 266. [Abstract] [Full Text] [PDF]

				J. E. Morley Editorial. Mobility Performance: A High-Tech Test for Geriatricians J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2003; 58(8): M712 - 714. [Full Text] [PDF]

				J. E. Morley, H. M. Perry III, and D. K. Miller Editorial: Something About Frailty J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2002; 57(11): M698 - 704. [Full Text] [PDF]

				J. E. Morley Editorial: Drugs, Aging, and the Future J. Gerontol. A Biol. Sci. Med. Sci., January 1, 2002; 57(1): M2 - 6. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation