Urinary incontinence is a common condition, associated with embarrassment, stigmatization, isolation, depression, and institutionalization.1 In addition to its medical, social, and public health consequences, the financial burden is substantial; estimated annual direct costs of incontinence in the United States were 16.3 billion dollars in 1995.2
Estrogen receptors have been demonstrated in the vagina, bladder, and pelvic floor. Estrogen increases the cell maturation index of the vagina, bladder trigone, and urethra;3 reverses microbiologic changes in the vaginal flora that occur after menopause;4 and elevates both urethral closure pressure5 and the sensory threshold of the bladder.6 However, small studies (including randomized controlled trials) of various estrogen preparations for treatment of urinary incontinence have reported mixed results,7–10 and a large randomized trial of postmenopausal women with coronary disease recently demonstrated that oral estrogen with progestin did not improve urinary function for those with incontinence.11 Even fewer data are available on the potential of hormone therapy to preserve urinary function in continent women, and no studies have specifically examined the variety of hormone regimens available. Although some epidemiologic investigations suggest an increase in incontinence for hormone users,12–16 these studies are generally cross-sectional, and it cannot be determined whether hormone use preceded development of incontinence or was prescribed after onset (ie, as a possible treatment). Thus, the purpose of this study was to prospectively assess the relation of postmenopausal hormone therapy, including duration and type of hormones, to development of urinary incontinence in 39,436 continent participants of the Nurses’ Health Study.
MATERIALS AND METHODS
The Nurses’ Health Study cohort was identified when 121,700 female, married registered nurses, aged 30–55 years, who resided in 1 of 11 U.S. states returned a mailed questionnaire in 1976. Since this time, these women have responded to biennial mailed questionnaires that request information on lifestyle factors and major medical diseases. The follow-up rate for the original cohort is more than 92%. Response to the mailed questionnaires is considered implied consent by participants; this study was approved by the Institutional Review Board of Brigham and Women's Hospital.
We identified information on menopausal status and use of postmenopausal hormone therapy, including duration, type of hormones, and dose of oral conjugated estrogen on each biennial questionnaire since 1976. In 1996, we requested information for the first time on incontinence, when we asked women “during the past 12 months, how often have you leaked urine or lost control of your urine?” with response categories of “never,” “less than once a month,” “once a month,” “two to three times a month,” “about once a week,” and “almost every day.” Women who responded that they lost urine were then asked “when you lose your urine, how much usually leaks?” with response categories of “a few drops,” “enough to wet your underwear,” “enough to wet your outer clothing,” and “enough to wet the floor.” These two items were repeated on the 2000 questionnaire. We defined women at risk of developing incontinence as those who reported never leaking or leaking less than once a month on the 1996 questionnaire. Incident cases of “occasional” incontinence from 1996 to 2000 were women who then reported leaking 1–3 times per month on the 2000 questionnaire, and incident cases of “frequent” incontinence were women who reported leaking at least once per week. Because associations between hormone therapy and incontinence were similar for occasional and for frequent incontinence, we combined the 2 in analyses. We separately analyzed a subgroup of more severe cases defined as frequent leaking of at least enough urine to wet the underwear (consistent with a validated severity index17).
The nurses’ responses to these questions were highly reliable. We conducted a reliability study by sending a second mailing several months after the return of the main Nurses’ Health Study questionnaire to a random sample of 200 women who reported various levels of urinary incontinence (n = 100) as well as no leaking (n = 100); 90% of the women remained within 1 category of their original response to the question on frequency of leaking, and 98% remained within 1 category of their original response to the item on quantity leaked.
For these analyses, we received data from 83,168 subjects because the questions on leaking urine were included only on initial mailing attempts in 1996 (the first 3 of 5 mailings). We excluded women who were premenopausal, who were missing data on postmenopausal hormone use in 1996 or were users of estrogen creams (these preparations were taken by only 1% of women in our population), or women who reported leaking urine once per month or more in 1996, leaving a total of 39,436 women who were considered at risk of developing incontinence from 1996 to 2000 and thus eligible for analysis. We classified women as postmenopausal in 1996 if they had reported natural menopause or hysterectomy with bilateral oophorectomy; women who underwent hysterectomy without bilateral oophorectomy were considered postmenopausal if they had reached the age at which natural menopause had occurred in 90% of the cohort (54 years for smokers and 56 for nonsmokers).
For the statistical analyses, we used logistic regression models to calculate odds ratios as an estimate of relative risks (RRs) and 95% confidence intervals (CIs) for the relation of postmenopausal hormone use to incident incontinence. Information on menopausal status, current hormone use, type of hormone, and dose were based on the information provided on the 1996 questionnaire, and duration of hormone therapy was determined by combining information from the 1976 through 1996 questionnaires. In the logistic regression models, we included the following potential confounding variables based on risk factors for incontinence previously identified in the scientific literature: age (continuous); hysterectomy (yes, no); cigarette smoking (never, current, past); body mass index (13 categories); race (Caucasian, African American, Asian American, Hispanic American); parity (nulliparous, 1 birth, 2 births, 3+ births); diabetes (yes, no); stroke (yes, no); and significant limitations in climbing 1 flight of stairs, walking 1 block, or bathing/dressing themselves (yes for any of the 3 items, no). We also adjusted for medications that could cause incontinence, including diuretics, calcium channel blockers, and angiotensin-converting enzyme inhibitors; however, because this had no effect on results, we did not include these in our final models. Information on potential confounders was updated through the baseline incontinence questionnaire in 1996. Tests of trend for ordered categories of time since last hormone use were calculated by including a continuous term in the multivariable models using the midpoint of each ordered category. Analyses were conducted using SAS statistical software, 8.2 (SAS Institute, Cary, NC).
Characteristics of hormone users were generally similar to those of women who had never taken hormone therapy (Table 1). However, women currently using postmenopausal hormones were slightly less likely to be smokers, had a somewhat lower prevalence of diabetes, and were younger at menopause. Hormone users were substantially more likely to have had a hysterectomy than were non-users.
Over the 4-year follow-up period, we identified 5,060 incident cases of occasional and 2,495 incident cases of frequent incontinence, for an average incidence rate of 3.2% per year and 1.6% per year, respectively. Overall, among women aged 50–55 years, the yearly incidence of incontinence was 3% for women who had never taken hormone therapy and rose to 3.8% in those who were current users of hormone therapy (Table 2). As expected, these yearly incidence rates rose steadily with increasing age; among those aged 71–75 years, the yearly incidence of incontinence was 5.3% for women who had never taken hormone therapy and 7.3% for current hormone users.
Compared with women who had never used hormones, those currently taking any hormone regimen were at significantly increased risk of developing urinary incontinence (defined as leaking at least once per month; Table 3). The age-adjusted RR of incontinence was 1.54 (95% CI 1.44, 1.65) for users of oral estrogen alone and 1.68 (95% CI 1.41, 2.00) for users of transdermal estrogen alone. These risks were slightly attenuated after controlling for potential confounders; there was a 45–60% higher rate of incident incontinence at least monthly for current compared to never users of hormone therapy, with comparable findings for oral or transdermal estrogen alone (RR 1.45, 95% CI 1.33, 1.57 and RR 1.55, 95% CI 1.29, 1.87, respectively) and for estrogen combined with progestin (RR 1.47, 95% CI 1.36, 1.59 for oral estrogen with progestin and RR 1.60, 95% CI 1.27, 2.02 for transdermal estrogen with progestin). Relative risks were generally somewhat stronger when we specifically considered women with severe incontinence (eg, for oral estrogen alone, multivariable-adjusted RR 2.02, 95% CI 1.45, 2.83 and for oral estrogen with progestin RR 1.78, 95% CI 1.50, 2.11; data not shown).
Because there was no apparent difference between estrogen alone or with progestin, we combined these 2 regimens in subsequent analyses. As suggested by a randomized trial of hormone therapy in women with incontinence in which negative effects were observed shortly after onset of use and persisted for 4 years of follow-up,11 we found no suggestion of a trend with duration (Table 3); the risk of incident incontinence was consistently elevated for shorter- and longer-term hormone users (eg, for oral therapy: RR 1.46, 95% CI 1.16, 1.84 for less than 1 year and RR 1.56, 95% CI 1.44, 1.69 for 10 or more years of use). Although likely the result of the smaller number of women using transdermal estrogen, results were not statistically significant for short-term users. There were also no significant differences in the effects of lower and higher doses of oral conjugated estrogen (Table 3).
However, for past hormone users (Table 4), there was little elevation in the rate of developing incontinence (RR 1.14, 95% CI 1.06, 1.23) compared with those who had never used hormones. This RR for past users was statistically significantly lower than the RRs for current users (RRs 1.45 to 1.60). Duration of past use did not appear to impact the risk of incontinence, but there was a highly significant trend of decreasing risk of developing incontinence with increasing time since last hormone use (P = .005); by 10 years after stopping hormone use, the risk was identical for past and never hormone users (RR 1.02, 95% CI 0.91, 1.14).
Finally, because we were particularly concerned about confounding by previous stroke or by physical limitations, we conducted analyses in which we excluded all subjects who had had a stroke or who reported significant limitations in their activities of daily living. However, results of these analyses were virtually identical to those reported above. In addition, because hysterectomy may be associated with increased incontinence,18 or possibly a decreased risk if accompanied by a bladder neck suspension, and was much more common in hormone users, we also conducted an analysis excluding women who had a hysterectomy (although 93% of the hysterectomies were performed many years before our baseline and thus likely would not affect our estimates of incident incontinence). Again, results were similar to the primary analyses (eg, for oral estrogen with progestin, RR 1.48, 95% CI 1.36, 1.60).
In this cohort of 39,436 female health professionals aged 50–75 years who reported no urinary incontinence, we found a significant increase in the risk of developing incontinence over 4 years in those taking postmenopausal hormone therapy at baseline. In the first study to examine several hormone regimens, we found that this risk was similar for users of estrogen alone and estrogen combined with progestin, as well as users of oral or transdermal estrogen. The elevated risk was evident for short-term and long-term current users but diminished after cessation of hormone therapy.
Several randomized trials have indicated that hormone therapy does not improve urinary symptoms,9–11 including trials of estrogen alone10 and with a progestin.9,11 Indeed, in the largest trial, among 1,525 women with coronary disease, the Heart and Estrogen/progestin Replacement Study11 found that women with incontinence assigned to combined estrogen with progestin were 45% more likely (P = .001) to experience a worsening of their condition compared with women given placebo; similar to our finding of an increased risk of developing incontinence beginning with short-term use, this worsening was noted at 4 months of follow-up and persisted over the 4 years of the trial.
Observational studies of primary prevention have noted an increase in incontinence for hormone users, although these have largely been cross-sectional studies13–16 in which it is not possible to distinguish the temporal relation between hormone therapy and incontinence (ie, many of the women may have started taking hormones subsequent to their incontinence), and no data were available on the hormone regimen or duration. In the only other prospective study we identified,12 among 382 women aged 20–59 years, there was also an elevated risk of incident leaking urine for women who reported taking hormone therapy at baseline (P = .01). In addition, studies of selective estrogen receptor modulators, provide indirect support for a harmful effect of estrogen therapy on incontinence. Levormeloxifene is a selective estrogen receptor modulator with estrogenic activity in the endometrium and vagina;19 in a randomized trial of 2,924 women assigned to levormeloxifene or placebo,19 there was 5-fold higher reporting of urinary incontinence over 10 months in women assigned to treatment than placebo (RR 4.99, 95% CI 3.55, 7.00) as well as approximately a 3-fold increase in uterovaginal prolapse (RR 3.44, 95% CI 2.13, 5.56). In contrast, in trials of raloxifene, a selective estrogen receptor modulator that appears to have limited estrogenic activity in the reproductive tract in animal models,20 the risk of surgery for urinary incontinence or pelvic organ prolapse was 50% lower (RR 0.50, 95% CI 0.31, 0.81) in treated women compared with those given placebo.21
A mechanism by which postmenopausal hormone therapy may cause urinary incontinence is not clear. In a small study,22 higher urinary estradiol levels were related to greater risk of pelvic organ prolapse; in combination with the data on prolapse discussed above for selective estrogen receptor modulators with different estrogenic profiles,19,21 these studies perhaps suggest that agents that are estrogen receptor agonists in the reproductive tract are detrimental to pelvic support and thus might lead to an increased likelihood of developing stress incontinence. In addition, postmenopausal hormone therapy has adverse neurologic effects23–25 and might also elevate risk of urge incontinence.
Several limitations of our study need to be considered. First, all information on hormone therapy and incontinence was self-reported. In a prospective study, any misclassification would bias results to the null; thus, we may have somewhat underestimated the apparent harm from hormone use. However, the bias is likely to be minimal because we believe the reports of hormone use are accurate in these health professionals. In addition, we found high reliability of responses to the items on urinary leaking. Moreover, in community-dwelling populations of similar age to our cohort, several investigators13,26 have established high validity of self-reported incontinence compared with a detailed physical examination. Second, in an observational study, it is possible that results are due to confounding. Our cohort of homogeneous health professionals likely reduces potential confounding because all subjects have good access to healthcare and sound health knowledge, regardless of their hormone use; nonetheless, confounding remains a possibility.
Finally, and most importantly, we did not have information on the type of incontinence (stress versus urge); thus, we cannot identify the etiology of the relation between hormone therapy and urinary incontinence. Still, our RR estimates represent the average effects of hormone therapy on stress and urge incontinence and have clear public health importance.
Overall, data from our study as well as others suggest that several hormonal therapies, including oral and transdermal estrogen alone or with progestin, may increase development of urinary incontinence, an addition to growing evidence of the potential clinical risks23–25,27,28 of postmenopausal hormone therapy.
1.Resnick NM. Urinary incontinence. Lancet 1995;346:94–9.
2.Wilson L, Brown JS, Shin GP, Luc KO, Subak LL. Annual direct cost of urinary incontinence. Obstet Gynecol 2001;98:398–406.
3.Thom DH, Brown JS. Reproductive and hormonal risk factors for urinary incontinence in later life: a review of the clinical and epidemiologic literature. J Am Geriatr Soc 1998;46:1411–7.
4.Molander U, Milsom I, Ekelund P, Mellstrom D, Eriksson O. Effect of oral oestriol on vaginal flora and cytology and urogenital symptoms in the post-menopause. Maturitas 1990;12:113–20.
5.Bhatia NN, Bergman A, Karram MM. Effects of estrogen on urethral function in women with urinary incontinence. Am J Obstet Gynecol 1989;160:176–81.
6.Fantl JA, Wyman JF, Anderson RL, Matt DW, Bump RC. Postmenopausal urinary incontinence: comparison between non-estrogen-supplemented and estrogen supplemented women. Obstet Gynecol 1988;71:823–8.
7.Fantl JA, Cardozo L, McClish DK. Estrogen therapy in the management of urinary incontinence in postmenopausal women: a meta-analysis. First Report of the Hormones and Urogenital Therapy Committee. Obstet Gynecol 1994;83:12–8.
8.Sultana CJ, Waters MD. Estrogen and urinary incontinence in women [review]. Maturitas 1994;20:129–38.
9.Fantl JA, Bump RC, Robinson D, McClish DK, Wyman JF. Efficacy of estrogen supplementation in the treatment of urinary incontinence. Obstet Gynecol 1996;88:745–9.
10.Jackson S, Shepherd A, Brookes S, Abrams P. The effect of oestrogen supplementation on post-menopausal urinary stress incontinence: a double-blind placebo-controlled trial. Br J Obstet Gynaecol 1999;106:711–8.
11.Grady D, Brown JS, Vittinghoff E, Applegate W, Varner E, Snyder T. Postmenopausal hormones and incontinence: the Heart and Estrogen/Progestin Replacement Study. Obstet Gynecol 2001;97:116–20.
12.Samuelsson EC, Victor FT, Svardsudd KF. Five-year incidence and remission rates of female urinary incontinence in a Swedish population less than 65 years old. Am J Obstet Gynecol 2000;183:568–74.
13.Diokno AC, Brock BM, Herzog AR, Bromberg J. Medical correlates of urinary incontinence in the elderly. Urology 1990;36:129–38.
14.Thom DH, van den Eeden SK, Brown JS. Evaluation of parturition and other reproductive variables as risk factors for urinary incontinence in later life. Obstet Gynecol 1997;90:983–9.
15.Brown JS, Grady D, Ouslander JG, Herzog AR, Varner RE, Posner SF. Prevalence of urinary incontinence and associated risk factors in postmenopausal women. Heart & Estrogen/Progestin Replacement Study (HERS) Research Group. Obstet Gynecol 1999;94:66–70.
16.Buchsbaum GM, Chin M, Glantz C, Guzick D. Prevalence of urinary incontinence and associated risk factors in a cohort of nuns. Obstet Gynecol 2002;100:226–9.
17.Sandvik H, Hunskaar S, Seim A, Hermstad R, Vanvik A, Bratt H. Validation of a severity index in female urinary incontinence and its implementation in an epidemiological survey. J Epidemiol Community Health 1993;47:497–9.
18.Brown JS, Sawaya G, Thom DH, Grady D. Hysterectomy and urinary incontinence: a systematic review. Lancet 2000;356:535–9.
19.Goldstein SR, Nanavati N. Adverse events that are associated with the selective estrogen receptor modulator levormeloxifene in an aborted phase III osteoporosis treatment study. Am J Obstet Gynecol 2002;187:521–7.
20.Black LJ, Sato M, Rowley ER, Magee DE, Bekele A, Williams DC, et al. Raloxifene (LY139481 HCI) prevents bone loss and reduces serum cholesterol without causing uterine hypertrophy in ovariectomized rats. J Clin Invest 1994;93:63–9.
21.Goldstein SR. An update on non-uterine gynaecological effects of raloxifene [review]. Eur J Cancer 2002;38(suppl 6):S65–6.
22.Bai SW, Jung BH, Chung BC, Kim SU, Kim JY, Rha KH, et al. Steroid hormone metabolism in women with pelvic organ prolapse. J Reprod Med 2002;47:303–8.
23.Grodstein F, Manson JE, Colditz GA, Willett WC, Speizer FE, Stampfer MJ. A prospective, observational study of postmenopausal hormone therapy and primary prevention of cardiovascular disease. Ann Intern Med 2000;133:933–41.
24.Wassertheil-Smoller S, Hendrix SL, Limacher M, Heiss G, Kooperberg C, Baird A, et al. Effect of estrogen plus progestin on stroke in postmenopausal women: the Women's Health Initiative: a randomized trial. JAMA 2003;289:2673–84.
25.Shumaker SA, Legault C, Rapp SR, Thal L, Wallace RB, Ockene JK, et al. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women's Health Initiative Memory Study: a randomized controlled trial. JAMA 2003;289:2651–62.
26.Herzog AR, Fultz NH. Prevalence and incidence of urinary incontinence in community-dwelling populations. J Am Geriatr Soc 1990;38:273–81.
27.Colditz GA, Hankinson SE, Hunter DJ, Willett WC, Manson JE, Stampfer MJ, et al. The use of estrogens and progestins and the risk of breast cancer in postmenopausal women. N Engl J Med 1995;332:1589–93.
28.Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA 2002;288:321–33.