Urinary incontinence is a common, debilitating, and costly problem in older women. Incontinence does not lead to death,1 but it causes substantial debility, social seclusion, psychologic stress, and economic burden.2,3 There have been few studies of risk factors for incontinence by type, although stress and urge types clearly differ in clinical presentation and response to treatment.4
We used baseline data from the Heart and Estrogen/progestin Replacement Study to determine prevalence of and risk factors for stress, urge, and mixed urinary incontinence among 2763 elderly, community-dwelling women with established coronary disease.
Materials and Methods
The study was a 4-year, multicenter, randomized, blinded trial on the effect of treatment with oral daily conjugated estrogens (0.625 mg) and medroxyprogesterone acetate (2.5 mg) on coronary disease events.5 Between January 1993 and September 1994, 2763 women at 20 clinical centers in the United States were enrolled. Postmenopausal women were eligible if they were younger than 80 years of age, had proven coronary heart disease, and had a uterus. Women were excluded if they had used sex hormones within the prior 3 months or were believed to be at risk for adverse effects of hormone therapy.6
At baseline, participants completed questionnaires on age, ethnicity, education, date of menopause, gravidity, parity, years since first childbirth, chronic illnesses (congestive heart failure, diabetes, hypertension), overall health status (excellent, good, fair, poor, or very poor), alcohol consumption (drinks per week), and smoking habits (current, former, or never and number of cigarettes per day among former and current smokers). Participants who rated their health statuses as fair, poor, or very poor were categorized as having poor overall health. They were asked to bring all prescription and over-the-counter medications to their visits, and those were recorded by research staff.
Participants also completed questionnaires on voiding habits and incontinence. The questions were modified from those used in previous epidemiologic studies.2,4,7 All questions began with “During the prior week, how many times, on average ….” Voiding questions included, “… have you had to go to the bathroom to urinate during the day?” and “… have you had to get up at night to go to the bathroom to urinate?” Responses were recorded as times per day or night. Incontinence questions included, “… have you unintentionally leaked some urine with coughing, sneezing, straining, laughing, or lifting?” (stress incontinence) and “… have you unintentionally leaked some urine before you could get to the bathroom?” (urge incontinence). Responses were recorded as times per week.
Participants who reported only stress losses were classified as stress incontinent, and those who reported only urge losses were classified as urge incontinent. Women with both types of incontinence who reported mostly stress losses were classified as stress-mixed incontinent, and those with mostly urge losses were classified as urge-mixed incontinent. Those who reported equal numbers of stress and urge losses were classified as mixed incontinent.
Physical examinations at baseline included measurements of height, weight, and waist and hip circumferences. Body mass index (BMI) was calculated as kg/m2, and waist-to-hip ratio was calculated as waist circumference divided by hip circumference.
Baseline characteristics of participants, diurnal frequency, nocturia, and prevalence of incontinence by type were reported as the mean with standard deviation or percentage. Mean diurnal, nocturnal, and number of weekly losses were compared using analysis of variance. To identify predictors of stress, urge, and mixed incontinence, in all analyses, women with no incontinence served as the reference group. There are no comparisons between women with different types of incontinence. Univariate logistic analyses with prevalence of type of weekly incontinence were used to evaluate variables associated with each type of incontinence. Those variables associated with weekly incontinence (P < 0.1) were included in multivariate, logistic regression models. Results are presented as three separate multivariate logistic models with odds ratios (OR) and 95% confidence intervals (CI) with outcomes as type of incontinence compared with continence. All analyses were done with SAS software (SAS Institute Inc., Cary, NC).
The mean (± standard deviation [SD]) age of the participants was 66.7 (±7) years (Table 1). Most participants were white (89%) or black (8%). Other characteristics of the cohort are displayed in Table 1.
Over half the participants reported urinary incontinence in the prior week (Table 2). On average, women reported 6 (±3) diurnal and 2 (±1) nocturnal voidings in 24 hours. Diurnal and nocturnal frequency did not differ between continent women and stress-incontinent women (Table 3). Women with urge and urge-mixed incontinence had more diurnal and nocturnal voidings than continent women (P < .001) or those with stress-incontinent (P < 0.001) women. Women with urge or stress incontinence had a mean of three incontinence episodes per week, while women with any type of mixed incontinence had a mean of eight episodes per week.
All variables in Table 1 that were significantly associated (P < 0.1) with stress, urge, or mixed incontinence were included in multivariate analyses. Table 4 presents the results of multivariate logistic-regression analyses. In multivariate models, white women had a threefold greater prevalence of stress incontinence than black women (OR 2.8, 95% CI 1.6, 5.0). Higher BMI (OR 1.1 per 5 units, 95% CI 1.0, 1.3) and higher waist-to-hip ratio (OR 1.2 for each 0.1 unit increase, 95% CI 1.0, 1.4) were also associated with increased prevalence of stress incontinence. White race, BMI and waist-to-hip ratio were predictors of stress-mixed incontinence (data not shown).
Urge incontinence was associated with advancing age (OR 1.2 for each 5 years, 95% CI 1.1, 1.3), diabetes (OR 1.5, 95% CI 1.1, 2.0), and urinary tract infections (OR 2.0 for at least two urinary tract infections in the past year). Those variables were also predictors of urge-mixed incontinence (data not shown).
The prevalence of weekly incontinence in this study was higher than previously reported,7 perhaps because incontinence was carefully measured and reported, or because the cohort was elderly and all had coronary heart disease. Weekly stress incontinence was present in 13% of the cohort, urge incontinence in 14%, and some form of mixed incontinence in 28%. Although the prevalence was higher, a similar predominance of mixed incontinence was reported in other studies.7
Among continent women, the frequency of diurnal voidings was 5.5 and nocturnal voidings 1.5, similar to frequencies reported in prior studies.8 Clinically, urinary frequency and nocturia are commonly reported by women with urge incontinence, but not by those with stress incontinence. Our findings supported this association, with urge-incontinent women reporting more frequent voiding than continent and stress-incontinent women. Voiding habits of women with stress incontinence did not differ from continent women.
Risk factors for urge and stress incontinence were different. The major predictors of urge incontinence were increasing age, diabetes, and urinary tract infections. The risk factors for urge incontinence were also predictors of mixed incontinence when the major component was urge. Other large population-based studies have reported age as an important risk factor for urge incontinence.7 The etiology of urge incontinence associated with aging is not clear, but might be multifactorial, including age-related lower urinary tract, vascular, and central nervous system changes affecting bladder control mechanisms.9
Our finding of 50% increased risk of urge incontinence in diabetic women was consistent with other large cross-sectional studies of women over the age of 60.2,4 The cause of bladder dysfunction and urinary incontinence in diabetics is not clear but might be an autonomic neuropathy affecting the innervation of the bladder,10 which might present in the early stages as urge incontinence. We also found a 50% increased risk of urge incontinence in women who reported two or more urinary tract infections in the prior year. Data from several studies showed urge incontinence lessens after antibiotic treatment for urinary tract infections.11 It was suggested that urinary tract infections stimulate detrusor hyperreflexia11 or that an Escherichia coli endo-toxin inhibits alpha-adrenergic contractions in the urethra, which reduces sphincter pressure and results in incontinence.12
The major predictors of stress urinary incontinence were white race, higher BMI, and higher waist-to-hip ratio, which were also predictors of mixed incontinence when the major component was stress. White women had nearly a threefold greater prevalence of stress incontinence than black women, which was noted previously,13 but the reason for the lower risk in black women is unclear. There might be genetically determined differences in anatomy or structural strength of the urethra and pelvic support structures that protect black women from stress incontinence. For example, black women were reported to have higher maximal urethral pressures, longer urethra, and elevated bladder necks than white women.13 Unfortunately, our cohort did not include adequate numbers of Hispanic or Asian women, or women of other ethnic groups to evaluate racial differences.
In our study, higher BMI was a risk factor for stress urinary incontinence. Other studies reported that obesity is associated with stress incontinence and that weight reduction can improve stress incontinence in morbidly obese women.14 Presumably, increasing body weight causes increased abdominal-wall weight, increased intra-abdominal pressure, and increased intra-vesicular pressure. We also found that higher waist-to-hip ratios were independent risk factors for stress incontinence even after adjusting for BMI. Higher waist-to-hip ratios indicated android or “apple” body shape, with weight distributed in the abdomen rather than the hips. Even at the same BMI, that fat distribution should increase abdominal wall pressure, intravesicular pressure, and risk for stress incontinence. If stress incontinence was primarily due to increased intra-abdominal pressure, women with increased waist-to-hip ratios (apples) might benefit from weight reduction, but other obese women (pears) might not.
In contrast to some studies,15 we did not find parity associated with incontinence. A similar lack of association with parity was found in other cohorts of older women.16 The difference in risk profiles in the various studies might indicate increased prevalence of chronic illnesses or predominance of urge incontinence among older women. As in other diseases, risk profiles often change with age.
Our study participants were volunteers with diagnosed coronary heart disease, so prevalence might not be similar in women without heart disease. However, we have no reason to believe that the risk factors that we identified would not be similar in other groups of women. We used the terminology of weekly incontinence, defined as women reporting one or more incontinent episodes in the prior week. It was possible that women who lost urine only once in the last year, but in the prior week, were classified as having weekly incontinence. However, that would likely change categories for only a few women. In this study, detailed questions modified from previous epidemiologic studies were used to define the types of clinical incontinence, and urinary diaries and urodynamic studies were not included. Incontinence symptomatology differed slightly but not significantly between prior studies using detailed questions and urinary diaries.17 Although lower correlation of urodynamic diagnoses and detailed questions were reported,18 measuring incontinence by detailed questions is more practical for a large epidemiologic study and might provide a more complete measure of the subject's actual experience over time in the subject's usual environment, rather than a single measure in a laboratory setting. For modifiable or preventable risk factors, trials to determine if risk reduction lowers incidence of incontinence should be conducted in women with stress or urge incontinence rather than in mixed populations.
1. Herzog AR, Diokno AC, Brown MB, Fultz NH, Goldstein NE. Urinary incontinence as a risk factor for mortality. J Am Geriatr Soc 1994;42:264–8.
2. Wetle T, Scherr P, Branch LG, Resnick NM, Harris T, Evans D, et al. Difficulty with holding urine among older persons in a geographically defined community: Prevalence and correlates. J Am Geriatr Soc 1995;43:349–55.
3. Fantl J, Newman D, Colling J, DeLancey J, Keeys C, McDowell B. Urinary incontinence in adults: Acute and chronic management. Clinical Practice Guideline, No. 2. Rockville, Maryland: AHCPR, 1996.
4. Diokno AC, Brock BM, Herzog AR, Bromberg J. Medical correlates of urinary incontinence in the elderly. Urology 1990;36:129–38.
5. Hulley S, Grady D, Bush T, Furberg C, Herrington C, Riggs B, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA 1998;280:605–13.
6. Grady D, Applegate W, Bush T, Furberg C, Riggs B, Hulley SB. Heart and estrogen/progestin replacement study (HERS): Design, methods, and baseline characteristics. Control Clin Trial 1998;19:314–35.
7. 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.
8. Burgio KL, Engel BT, Locher JL. Normative patterns of diurnal urination across six age decades. J Urol 1991;145:728–31.
9. Artibani W. Diagnosis and significance of idiopathic overactive bladder. Urology 1997;50:25–32.
10. Andersen JT, Bradley WE. Abnormalities of bladder innervation in diabetes mellitus. Urology 1976;7:442–8.
11. Bhatia NN, Bergman A. Cystometry: Unstable bladder and urinary tract infection. Br J Urol 1986;58:134–7.
12. Nergardh A, Boreus LO, Holme T. The inhibitory effect of coli-endotoxin on alpha-adrenergic receptor functions in the lower urinary tract. An in vitro study in cats. Scand J Urol Nephrol 1977;11:219–24.
13. Mallett VT, Bump RC. The epidemiology of female pelvic floor dysfunction. Curr Opin Obstet Gynecol 1994;6:308–12.
14. Deitel M, Stone E, Kassam HA, Wilk EJ, Sutherland DJ. Gynecologic-obstetric changes after loss of massive excess weight following bariatric surgery. J Am Coll Nutr 1988;7:147–53.
15. Thom D, Brown JS. Reproductive and hormonal risk factors for urinary incontinence: Review of the clinical and epidemiologic literature. J Am Geriatr Soc 1998;46:1411–7.
16. Diokno A, Brown M, Brock B, Herzog A, Normolle D. Clinical and cystometric characteristics of continent and incontinent noninstitutionalized elderly. J Urol 1988;140:567–71.
17. Wyman JF, Choi SC, Harkins SW, Wilson MS, Fantl JA. The urinary diary in evaluation of incontinent women: A test-retest analysis. Obstet Gynecol 1988;71:812–7.
18. Kirschner-Hermanns, R, Scherr PA, Branch LG, Wetle T, Resnick NM. Accuracy of survey questions for geriatric urinary incontinence. J Urol 1998;159:1903–8.