Huang AJ, Brown JS, Kanaya AM, Creasman JM, Ragins AI, Van Den Eeden SK, Thom DH. Quality-of-Life Impact and Treatment of Urinary Incontinence in Ethnically Diverse Older Women. Arch Intern Med. 2006;166(18):2000–2006. doi:10.1001/archinte.166.18.2000
To identify the factors associated with greater quality-of-life impact, treatment seeking, and use of treatments for urinary incontinence in ethnically diverse older women.
Cross-sectional analysis of a population-based cohort of 2109 middle-aged and older women who were randomly selected from age and race/ethnicity strata. Data were collected by self-report questionnaires and in-person interviews. Multivariable logistic regression was used to identify predictors of high quality-of-life impact (Incontinence Impact Questionnaire [IIQ] score ≥75th percentile), treatment seeking, and use of treatments for incontinence.
More than one fourth (n = 603) of the study participants (including 96 black [16%], 123 Latina [20%], 65 Asian [11%], and 309 white [51%] women) reported weekly incontinence. After clinical severity and other factors were adjusted for, women were more likely to experience high quality-of-life impact if they had nighttime incontinence (odds ratio [OR], 2.5; 95% confidence interval [CI], 1.3-4.9), coital incontinence (OR, 1.9; 95% CI, 1.1-3.3), or comorbid fecal incontinence (OR, 2.2; 95% CI, 1.2-4.2). Predictors of treatment seeking included older age (OR, 1.6 per 10 years; 95% CI, 1.2-2.0); higher IIQ score (OR, 4.6 for highest IIQ quartile vs lowest IIQ quartile; 95% CI, 2.5-8.4), and higher household income (OR, 2.6 for income ≥$100 000/y vs <$20 000/y; 95% CI, 1.0-2.7).
Clinicians seeking to evaluate the impact of incontinence on women's lives should assess not only the clinical severity of their symptoms but also the specific context in which symptoms occur. The prevalence of treatment seeking for incontinence is low across all ethnic groups, even when women have clinically severe symptoms and access to a health provider.
Urinary incontinence is a common problem in older women, with an estimated prevalence of 41% to 57% among women older than 40 years in the United States.1- 3 Among older community-dwelling women, incontinence is associated with social isolation, depression, decreased physical activity, falls and fractures, and nursing home admission.4- 8 Despite growing concern about the underdiagnosis of incontinence, our understanding of the quality-of-life impact and determinants of treatment seeking for this problem is limited.
Within the United States, the existing research on quality-of-life impact and treatment seeking has focused almost exclusively on white women. As a result, we do not know whether determinants of impact or treatment identified in white women are generalizable to more diverse populations.9 Within individual studies, there has also been considerable variation in participants' insurance status, distance to a medical facility, and access to a health care provider.10,11 There are few data to indicate whether factors that are seemingly associated with seeking treatment for incontinence may simply be reflective of more general disparities in access to medical care.
Furthermore, while we know only a little about patterns of treatment seeking, we know even less about the actual treatments and self-care practices that are used by incontinent women in the United States. There are very few population-based data to indicate whether women are more likely to use behavioral vs pharmacologic treatments, for example, or to rely on diapers and absorbents in lieu of prescribed medical therapies for incontinence.
In our study, we examined the characteristics that are associated with greater quality-of-life impact and treatment seeking for urinary incontinence in an ethnically diverse population of older adult women enrolled in a prepaid health plan, all of whom reported weekly ongoing incontinence symptoms. We also examined the spectrum of treatments (ie, behavioral, pharmacologic, and surgical) and coping strategies used by these women to manage their symptoms.
We performed a cross-sectional analysis of baseline data from the Reproductive Risks of Incontinence Study at Kaiser (RRISK), a cohort of 2109 women enrolled in the Kaiser Permanente Medical Care Program of Northern California, an integrated health care delivery system serving approximately 30% of the northern California population. To be eligible, women had to be between 40 and 69 years of age on January 1, 1999; to have been continuously enrolled in a Kaiser Permanente clinic since the age of 18 years; and to have given birth to at least half of their children within the Kaiser Permanente system. Details on the sampling process used to construct this cohort with a race/ethnicity composition of 20% black, 20% Latina, 20% Asian, and 40% white have been previously described.12 To focus our analysis on those women whose symptoms were clinically significant enough to warrant evaluation and/or treatment, we identified the 603 participants who reported at least weekly incontinence.
Demographic characteristics, medical and surgical histories, incontinence symptoms, quality-of-life impact, and use of treatment strategies were assessed by self-report questionnaires as well as by in-person interviews. The majority of interviews were conducted in the woman's home (57%) or at a Kaiser Permanente clinic (28%). Interviewers made a specific point of reviewing and clarifying participants' answers to questions about incontinence.
We asked women to self-identify their race/ethnicity as non-Latina white/Caucasian, Hispanic/Latina, African American/black, or Asian American/Asian. A bilingual interviewer was available for women who preferred to be interviewed in Spanish, but this option was selected by fewer than 1% of the participants. Questionnaires were available in English only. Informed consent was obtained by telephone and in writing at the time of data collection.
Details on the specific measures used to assess the frequency and severity of women's incontinence symptoms have been described elsewhere.12 Women were asked, “During the past 12 months, on average, how often have you leaked urine, even a small amount?” We asked women with at least weekly incontinence to recall the number of episodes in the past 7 days occurring “with an activity like coughing, lifting, sneezing, or exercise” to distinguish stress incontinence and the number of episodes in the past 7 days occurring “with a physical sense of urgency” to distinguish urge. We classified women as having stress incontinence if they reported only or predominantly stress episodes in the past 7 days and as having urge incontinence if they reported only or predominantly urge episodes in the past 7 days. Incontinence not associated with either an activity or a sense of urgency was characterized as “other urinary incontinence.” Women were also asked how old they were when they first began leaking urine.
Clinical severity of incontinence was determined based on incontinence frequency and amount of urine lost per episode, using the validated Sandvik Severity Scale.13 Quality-of-life impact of incontinence was measured using the Incontinence Impact Questionnaire (IIQ), which assesses impact in 4 separate domains (physical activity, emotional health, social relationships, and travel) to assign an overall impact score.14 One item from the physical domain (“Has urine leakage affected your physical health?”) was not included to avoid duplication with other questions elsewhere on the RRISK survey. In computing the physical activity subscale and overall IIQ scores, we imputed the average of the other 5 items in the physical activity subscale for this missing item. As recommended by the authors of this instrument, the IIQ was scored from 0 to 400, with higher scores indicating greater impact on daily life.14 For purposes of analysis, we designated women who had a total IIQ score greater than or equal to the 75th percentile (IIQ score ≥53) for our cohort as having high quality-of-life impact.
Women were shown a list of possible treatments for incontinence and asked to indicate all measures that they were currently taking to help them with their symptoms. We classified timed urination and pelvic muscle training as “behavioral” treatment for incontinence. Medications that were Food and Drug Administration–approved exclusively for the treatment of incontinence, such as oxybutynin and tolterodine, were categorized as “pharmacologic” treatment for incontinence. Other medications that could be prescribed for other indications were not included in this category. We considered women's use of diapers or absorbents, restriction of fluid intake, and avoidance of activities that might provoke incontinence to be “coping strategies” for incontinence.
We used backward-elimination multivariate logistic regression analysis to identify characteristics associated with experiencing high quality-of-life impact as a result of incontinence, seeking treatment for incontinence, and using specific medical therapies or coping strategies for incontinence. Age and race/ethnicity were included in all multivariate models a priori. Other variables were initially entered into a given model only if they were found to be associated with the outcome in question at a significance of P<.20 in bivariate analysis. Variables were retained in the model if they remained associated at P<.20 in multivariate analysis. All analyses were performed with SAS software (Version 9.1; SAS Institute, Cary, NC).
The baseline characteristics of the 603 women with weekly incontinence are shown in Table 1. The mean ± SD age of participants was 57 ± 9 years. Nearly 50% of the women were nonwhite, and more than 50% had clinically severe incontinence according to the Sandvik Severity Scale.
The mean ± SD IIQ score for participants was 40 ± 58, and the median IIQ score was 17 (interquartile range, 4-53). High quality-of-life impact (IIQ scores ≥53 [75th percentile]) was associated with Latina ethnicity, lower household income, lower educational status, lack of employment, poorer overall health status, greater use of medications, greater clinical severity of incontinence, nighttime incontinence, mixed or other type of incontinence, coital incontinence, comorbid fecal incontinence, and younger age at onset of incontinence in univariate analysis (Table 1). After race/ethnicity and clinical incontinence severity were controlled for in multivariate analysis, women were more likely to report high quality-of-life impact if they had clinically severe incontinence, nighttime incontinence, coital incontinence (incontinence with sexual activity), comorbid fecal incontinence, lower education level, or poorer health status (Table 2). Age, income, employment status, medication use, and clinical type of incontinence were not significantly associated with high quality-of-life impact.
Only 42% of women with weekly incontinence reported seeing a physician for their symptoms. Although the proportion of women who had sought treatment was lower for blacks (33%) and Asians (34%), compared with whites (42%), this difference was not statistically significant (P = .08 and P = .14, respectively). The proportion of Latina women who reported seeking treatment (44%) was similar to that of white women. In multivariate analysis, increasing IIQ score was a strong independent predictor of consulting a physician (P<.01 for trend) (Table 3). Other characteristics associated with treatment seeking were increasing age and higher household income (P<.01 for trend for both). Race/ethnicity, education, employment status, timing of incontinence, age at onset of incontinence, coital incontinence, clinical type of incontinence, and clinical severity did not predict treatment seeking.
Behavioral treatments were the most widely used types of treatments, with 30% of women practicing pelvic floor exercises and 12% practicing timed urination (Table 4). Forty percent of women using behavioral treatments indicated that they had never sought treatment from a physician, suggesting that they learned about these therapies from another source. Only 5% of women reported taking medications such as oxybutynin or tolterodine, and only 4% reported undergoing surgery for incontinence.
In multivariate analysis, there was a strong association between using behavioral treatments and increasing IIQ score (P<.01 for trend for all behavioral treatments). Women with IIQ scores in the highest quartile were more than 3 times as likely to use pelvic floor exercises (odds ratio [OR], 3.2; 95% confidence interval [CI], 1.9-5.5) and more than 2 times as likely to use timed urination (OR, 2.7; 95% CI, 1.2-6.4) than women with IIQ scores in the lowest quartile. Women were also more likely to practice timed urination if they were college graduates (OR, 1.9; 95% CI, 1.1-3.3) or if they had clinically severe incontinence (OR, 2.3; 95% CI, 1.3-4.2). Women were more likely to use pelvic floor exercises if they had predominantly stress incontinence (OR, 2.0; 95% CI, 1.2-3.1) rather than urge incontinence. Incontinence-specific medications were used more often by women with higher IIQ scores (P<.01 for trend across quartiles) and by women with poorer overall health status (P<.01 for trend).
The majority of women were practicing 1 or more coping strategies for incontinence (Table 4). Women who sought treatment from a physician were more likely to report using diapers, pads, or absorbents; restricting travel; practicing frequent urination; or avoiding activities that might cause incontinence (OR, 3.3; 95% CI, 2.0-5.5 for any coping strategy). In multivariate regression analysis, women were more likely to use diapers and absorbents if they had a higher IIQ score (OR, 2.7; 95% CI, 1.5-4.8 for highest quartile vs lowest quartile; P<.01 for overall trend), a higher household income (OR, 3.7; 95% CI, 1.5-9.6 for ≥$100 000/y vs <$20 000/y; P<.01 for overall trend), predominantly stress rather than urge incontinence (OR, 2.0; 95% CI, 1.3-3.3), or used a greater number of medications (P<.01 for trend). Women were more likely to have frequent urination if they were (1) older (OR, 1.3; 95% CI, 1.1-1.7 per 10-year interval), (2) college graduates (OR, 2.2; 95% CI, 1.5-3.2), or (3) sexually active, whether or not they experienced coital incontinence (OR, 2.3; 95% CI, 1.4-3.8 for women with coital incontinence; OR, 2.1; 95% CI, 1.3-3.8 for women without coital incontinence).
This study reveals several important trends in the quality-of-life impact and treatment of urinary incontinence in ethnically diverse women in the United States. In this population of women with at least weekly ongoing incontinence, quality-of-life impact varied significantly depending on factors related to the timing and setting of symptoms. After clinical incontinence severity and other factors were controlled for, women were more likely to have impact scores in the highest quartile if they had nighttime incontinence, coital incontinence, or comorbid fecal incontinence. These findings suggest that the particular context in which women's incontinence occurs may be as important as the clinical severity of their symptoms in determining quality-of-life impact. As a result, current approaches to evaluating incontinence in general practice, which emphasize identification of the clinical type of incontinence and screening for modifiable risk factors,15 may fail to identify those factors that determine whether incontinence is truly disruptive to women's lives.
Fewer than 50% of the women in this analysis reported seeking treatment from a physician despite having clinically frequent incontinence. Although the prevalence of treatment seeking for moderate to severe incontinence in other research has ranged from 28% to 45%, not all women in these studies had access to a health care provider.10,11,16,17 In contrast, all of the women participating in our study were insured and had continuous access to either a primary care provider or a gynecologist at a Kaiser Permanente clinic since the age of 18 years. These results suggest that health system access factors alone cannot account for the low rates of treatment seeking for incontinence that have been observed in the community.
After adjustment for clinical severity and quality-of-life impact of incontinence, age remained a significant predictor of treatment seeking in our cohort. One explanation for this finding is that because of comorbidities that require ongoing care, elderly women may have more frequent interactions with health care providers than younger women. Nevertheless, we did not find a significant association between treatment seeking and health status, and the association with age persisted after adjustment for health care use factors such as number of medications. These observations suggest that older age may be predictive of treatment seeking independent of other characteristics that might be expected to predict more frequent interaction with providers.
Women with higher household incomes in our cohort were more likely to have sought treatment for incontinence independent of quality-of-life impact and other factors. This association was primarily significant for women in the highest income category (≥$100 000/y). Very few previous studies of incontinence have collected data on household income, and we are not aware of any that have reported a positive association between income and treatment seeking.
Although there was a trend toward less treatment seeking among blacks and Asians, we did not find race/ethnicity to be a major independent determinant of treatment seeking. This finding is consistent with a recent survey of middle-aged women from the Study of Women Across the Nation, in which no significant differences across racial groups were detected after differences in other clinical and demographic characteristics were taken into account.18 In a telephone survey of older women in an urban academic practice, no significant differences in treatment seeking were noted between African American and white women.19 In light of this, our research suggests that race/ethnicity may not play as great a role as other demographic characteristics in determining whether women seek treatment for incontinence.
We did not ask participants to explain why they did or did not seek treatment from a health care provider. Reasons for failing to seek treatment that have been cited in previous research include a lack of knowledge about incontinence, a misperception that it is a normal part of aging or childbirth, and a belief that symptoms are inappropriate for medical intervention.16,20- 22 However, these barriers were identified predominantly among white women, and it is possible that different factors may influence women from other ethnic or cultural backgrounds.
Women with moderate to severe urinary incontinence in our study were more likely to be using 1 or more coping strategies than a standard medical treatment for this problem. This finding is consistent with a recent cross-sectional survey of predominantly white US women aged 18 years and older, in which women with self-reported incontinence in the previous 30 days were more likely to be using pads or absorbents, practicing frequent urination, or restricting their activities than taking medications, using pelvic floor exercises, or undergoing surgery for incontinence.23 One possible explanation is that women may turn to coping strategies after finding that standard medical therapies either fail to treat their symptoms or cause adverse effects. Given that only a minority of women in our study reported ever consulting a physician about incontinence, however, it is also likely that many women may regard incontinence as more of a personal self-care or hygiene problem as opposed to a true medical condition.
The mean IIQ score in this cohort of generally well-functioning, community-dwelling women was 40 (out of a possible 400), suggesting that the overall impact of incontinence in these women was low. While studies of women recruited from incontinence-specific clinics or undergoing surgery for incontinence have generally reported higher IIQ scores,24- 27 other studies involving community-based populations have reported even lower mean IIQ scores of 4.3 to 28.0.28- 31 Further research may help demonstrate whether factors associated with higher quality-of-life impact in this population are generalizable to populations in which the overall impact of incontinence is higher.
Although our study has multiple strengths, including its use of a population-based sample, careful measures of incontinence severity, and detailed assessment of incontinence management strategies, several limitations should be noted. First, data on consultation and treatment patterns were collected by self-report; we did not attempt to get correlation of responses by examining pharmacy databases or physician records. Furthermore, we asked women only about current symptoms and treatment and, as a result, could not determine whether women may have sought treatment and been cured of their incontinence or may have previously tried conventional therapies before turning to self-care practices. Finally, our study did not attempt to differentiate between different ethnic subgroups within race (eg, East Asian, Southeast Asian, Indian Subcontinent, Central American, South American, or Mexican). Further research is needed to determine if these trends in impact or treatment differ by ethnic subgroup.
In summary, our findings suggest that health care providers seeking to evaluate the impact of urinary incontinence on women's lives should not only assess the clinical severity and type of their urine leakage but also ask about the specific context in which these symptoms occur. After adjustment for other demographic and clinical factors, only quality-of-life impact, older age, and higher household income are significant predictors of treatment seeking for this problem.
Correspondence: Alison J. Huang, MD, MPhil, General Internal Medicine Section, San Francisco Veterans Affairs Medical Center, 4150 Clement St, Bldg 18, VAMC 111A1, San Francisco, CA 94121 (email@example.com).
Accepted for Publication: June 8, 2006.
Author Contributions: Dr Huang had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Huang, Brown, Van Den Eeden, and Thom. Acquisition of data: Brown, Ragins, Van Den Eeden, and Thom. Analysis and interpretation of data: Huang, Brown, Kanaya, Creasman, Van Den Eeden, and Thom. Drafting of the manuscript: Huang, Brown, Ragins, and Thom. Critical revision of the manuscript for important intellectual content: Huang, Brown, Kanaya, Creasman, Van Den Eeden, and Thom. Statistical analysis: Huang, Creasman, and Thom. Obtained funding: Brown, Ragins, Van Den Eeden, and Thom. Administrative, technical, and material support: Brown, Ragins, and Van Den Eeden. Study supervision: Brown, Van Den Eeden, and Thom.
Financial Disclosure: None reported.
Funding/Support: This study was funded by National Institute of Diabetes and Digestive and Kidney Diseases grant DK53335 and by National Institute of Diabetes and Digestive and Kidney Diseases–Office of Research on Women's Health Specialized Center of Research grant P50 DK064538.