The prevalence of lower urinary tract symptoms increases with advancing age and particularly by the time a women enters the perimenopausal years and into menopause.1–3 However, it remains unclear whether the increasing prevalence of urinary tract symptoms is due to the hormonal changes that occur during the menopausal transition or advancing age. It has been shown that a hypoestrogenic state is associated with thinning of the urethral mucosa, changes in urethral closure pressure from loss of urethral sphincter tone, and alteration of the urethrovesical angle.4 These physiologic changes that occur in the urogenital tissue are factors that have been shown to affect the severity of urinary incontinence. Studies5–7 exploring the relationship between hormonal status and incontinence have been limited by the inadequate assessment of E2 levels, menopausal status, and severity of lower urinary tract symptoms, including incontinence. Our aim was to estimate the relationship between changes in E2 levels over time and lower urinary tract symptoms in premenopausal women as they transition to the different stages of menopause. We hypothesized that declining E2 levels during the perimenopause transition to menopause correlates with an increase of lower urinary tract symptoms and, more specifically, incontinence.
MATERIALS AND METHODS
The proposed investigation was conducted as part of the Penn Ovarian Aging Study. The Penn Ovarian Aging Study is a prospective, longitudinal cohort of women enrolled to evaluate the relationship between changes in reproductive hormone levels and various symptoms associated with the menopausal transition. Women were recruited from Philadelphia County from 1996–1997 and identified by random digit dialing. Women were initially enrolled between the ages of 35 and 47 years if they had menstrual cycles in the normal range (22–35 days) and an intact uterus with at least one ovary. Exclusion criteria included use of current hormonal medications, oral contraceptive pills within the last 3 months, and use of psychotropic medications. Women were also excluded if there was a history of alcohol or drug abuse in the past year, pregnancy or breastfeeding, a serious medical problem that could compromise ovarian function, or major psychiatric disorder that would interfere with their ability to complete the study. In this cohort, none of the women had a history of prior incontinence surgery. Written informed consent was obtained from all subjects and the institutional review board at the University of Pennsylvania approved the study.
As part of this study, subjects were assessed initially at 8-month intervals until assessment period 6 and then yearly for a total of 11 assessments over 10 years. At each assessment, subjects were seen on two separate visits 1 month apart during the first 6 days of the menstrual cycle. Blood specimens, anthropometric measures, and extensive questionnaires were obtained from the subjects at each visit. Repeat blood samples one month apart were collected to ensure reliable point estimates of hormone measurements. Estradiol hormone levels, body mass index, and psychosocial and demographic variables were measured at each visit. At the last assessment period, the Bristol Female Lower Urinary Tract Symptoms Scored Form (BFLUTS-SF) questionnaire was also administered to the participants. Using the combination of previously collected and prospectively collected measures, there will be 10 years of E2 hormone measurements and recent data on lower urinary tract symptoms to analyze.
Blood samples were centrifuged and frozen in aliquots at –80°C. Assays were conducted in the laboratory of the General Clinical Research Center of the Hospital of the University of Pennsylvania. Estradiol levels were measured by radioimmunoassay using Coat-A-Count commercial kits (Diagnostic Products, Los Angeles, CA). Estradiol assays were timed to the subject’s early follicular phase, repeat measurements were taken to ensure reliable point estimates, and assays were repeated if values differed by more than 15%. The intraassay and interassay coefficients of variation were consistently less than 5% for all hormones measured.
Trained research interviewers conducted the interviews in person with the study participants and performed the data collection. Self-report questionnaires were also used. In the latest assessment period, the BFLUTS-SF was added to the various self-report questionnaires given to the study population to investigate the association between lower urinary tract symptoms and E2 change through the menopausal transition. The BFLUTS-SF8–9 questionnaire has been shown to be a valid and reliable instrument that can be used to ascertain a woman’s lower urinary tract symptoms and their effect on her quality of life and sexual function. This questionnaire is a 19-item, multidimensional self-report instrument assessing key dimensions of lower urinary tract symptoms, sexual function, and quality of life. There are five subscales within the questionnaire. Three subscales focus on incontinence symptoms (BFLUTS-IS), voiding symptoms (BFLUTS-VS), and filling symptoms (BFLUTS-FS). The remaining two focus on sexual function (BFLUTS-sex) and effect on a woman’s quality of life (BFLUTS-Qol). Responses range from none to always, no problem to serious problem, and “0” to “4,” with 4 representing the most severe symptom. No cutoff values have been established in the literature; therefore, the domain scores were analyzed as continuous measures. Subscore changes have been shown to be more sensitive than a single combined score for lower urinary tract symptoms.9 For our analysis we focused on the domain scores (cumulative summary score of all questions in a particular domain) specifically addressing lower urinary tract symptoms (BFLUTS-IS, BFLUTS-VS, BLUTS-FS).
At each assessment period, the participant’s current menstrual dates and two prior menstrual periods were ascertained by interview and by each woman’s daily symptom diary. With the menstrual information given, each woman was assigned a hormonal status based on the Stages of Reproductive Aging Workshop consensus statement.10 The stages are as follows: Premenopausal, regular menstrual cycles in the 22–35 day range; early transition, change in cycle length of 7 days or longer in either direction from the participant’s own baseline for at least 2 cycles; late transition, 3 to 11 months of amenorrhea; and menopause, 12 months or more of amenorrhea without hysterectomy.
Anxiety was assessed using the Zung Anxiety Index.11 Established score ranges for the Zung Anxiety Index include normal anxiety, (20–35) moderate anxiety (36–47), and high anxiety (48–60). Demographic information including, race, parity, history of hysterectomy, tobacco use, and body mass index was also obtained. Means and standard deviation of E2 levels were calculated. Because of the skewed distributions, hormones values were analyzed on the log scale and reported as geometric means.
We explored the association between the changes in E2 levels that occurred from the starting point of entry for each woman in the Penn Ovarian Aging cohort to the current assessment period and the lower urinary tract symptoms domains of the BFLUTS-SF. Participants were categorized by the direction of change of hormone levels over time which reflects the transition into menopause. Values obtained through the 11 assessment periods, over 10 years of evaluation (a maximum of 22 hormone values per woman) were then plotted. A regression coefficient (slope) corresponding to the overall trend of change in hormone values was obtained. These regression coefficients were used to summarize each participant’s hormonal trend into four different E2 quartiles. The four E2 quartiles were: increase, slight decrease, sharp decrease, and no significant change. The distribution of each participant’s current menopausal stage by E2 quartiles was compared as well as the BFLUTS-SF domain scores. A univariable analysis was performed to assess the association between BFLUTS-SF domain scores and known risk factors (age, race, body mass index (BMI), parity, anxiety, smoking, and history of prior hysterectomy) of lower urinary tract symptoms using nonparametric testing such as Kruskal-Wallis or the rank sum test. A multivariable linear regression model was used to estimate the association between trend of E2 change through the menopausal transition and urinary symptoms, adjusting for potential confounders. A variable was considered a confounder, and included in the model, if the estimated associations of interest change by 15% or if the variable is associated with the outcome (BFLUTS-SF subscale scores) with a P<.2.12 Data analysis utilized Stata 9.2 software (StataCorp LP, College Station, TX), with a two-tailed interpretation of tests and a P value less than 0.05 considered statistically significant.
We defined a clinically meaningful difference in BFLUTS-SF’s subdomain scores as 0.5 standard deviation that corresponded to a difference in severity of one level on a single question in the BFLUTS-SF questionnaire. Sample size calculations were performed using a two-group comparison of the difference in mean BFLUTS-SF incontinence domain scores between two of the four E2 quartile groups. We determined at a two-tailed alpha error of 0.05, power 80%, and a mean incontinence subscale score of 2.66±2.81 in our cohort that 64 patients will be needed in each group of categorized overall E2 change to detect a difference of 0.5 standard deviation in mean BFLUTS-SF incontinence subscale scores.
Of the 436 women originally enrolled in the Penn Ovarian Aging Study, 311 participated at assessment period 11. During that assessment, 300 women completed the BFLUTS-SF questionnaire and therefore were eligible for inclusion into this substudy. Demographic characteristics of the study cohort and mean domain scores are displayed in Table 1. Overall, the mean (±standard deviation) BFLUTS-SF score for the study participants was 8.38 (±2.81). The mean age of the cohort was 50.6 (±3.52) years with a mean parity of 3.2 (±2.1). Forty-six percent of the women in this cohort had stress incontinence.
Univariable associations between risk factors and the continuous outcome incontinence, filling, and voiding dysfunction scores are displayed in Table 2. African-American women had higher scores in all three domains in comparison with white women and had a statistically significant increase in filling symptoms (P=.002). Smokers had a significantly higher urinary filling symptoms score (P=.001). Menopausal stage and age were not associated with the three BFLUTS-SF domains. Women with a BMI more than 35 reported an increase in incontinence (P=.002) and urinary filling (P=.008) symptoms. Women with moderate and high anxiety had a statistically significant increase in lower urinary tract symptoms scores through all three domains of the BFLUTS-SF. No correlation was seen between E2 levels at assessment period 11 and the incontinence scores (ρ=0.06), filling scores (ρ=0.04) and voiding scores (ρ=0.03). A moderate correlation was seen between parity and incontinence scores (ρ=0.20), whereas a weak correlation was seen between parity and urinary filling (ρ=0.15) symptoms and voiding (ρ=0.14) symptoms. The E2 quartiles were not associated with urinary filling and voiding dysfunction scores. However, women with decreasing E2 quartiles had lower urinary incontinence symptom scores that approached significance (P=.09).
The regression coefficients describing each participant’s E2 trend was categorized based on quartiles into four categories: sharp decrease (n=75), slight decrease (n=75), no change (n=75), and increase (n=75). These groups characterized the change in measured E2 levels observed over the 10-year study period and was reflective of their current menopausal status (P<.001). Seventy-three percent of the women in the sharp decrease group were in menopause. The women in the slight decrease category were almost equally distributed between early transition (33%), late transition (30%), and menopause (37%). The majority of women in the no change group were in the early transition (46%), followed by late transition (22%), menopause (21%), and premenopause (11%). The distribution of women in the increase group was similar to the no change group, with the majority of women in early transition (51%), followed by menopause (27%) and late transition (15%). Because the changes in E2 over time as represented by the E2 quartiles and current menopausal stage were collinear, menopausal stage was not included in the final linear regression model.
Table 3 represents the multivariable models assessing the association between the trend of change in longitudinally measured E2 levels and the three BFLUTS-SF domains evaluated at assessment 11. Women with a sharp decline in E2 over the 9 years had significantly lower incontinence scores (–0.93, 95% confidence interval [CI] –1.8 to –0.02) compared with women in the “no change” E2 quartile group. This decrease represents a change in severity by one level on a single question in the incontinence domain of the BFLUTS-SF questionnaire. The “slight decrease” and “increase E2” quartile groups were not associated with incontinence scores. Women between the ages of 45 to 49 years had a significant increase in incontinence score compared with women aged younger than 45 years after adjusting for changes of E2 through the menopausal transition. To further determine if a specific form of urinary incontinence was associated with the E2 quartiles, the final models were repeated using the individual question pertaining to stress incontinence and urge incontinence within the BFLUTS-SF incontinence domain as the outcome. We found that women in the sharp decrease E2 quartile had a significant reduction in stress incontinence symptoms (–0.48, 95% CI –0.80 to –0.16, P=.003). Urge incontinence was not associated with the change in E2 levels through the menopausal transition. When examining the relationship between E2 quartiles and the remaining BFLUTS-SF domain scores (urinary filling symptoms and voiding dysfunction), no significant change in domain scores was seen. High anxiety was associated with worsening scores in all three domains (incontinence, filling, voiding).
Our goal was to identify risk factors for lower urinary tract symptoms and to determine if the change in E2 level through the menopausal transition significantly affected a woman’s lower urinary tract symptoms. Unexpectedly, we found that women with a sharp decline in E2 levels through the menopausal transition have significantly decreased urinary incontinence symptoms. Furthermore, we found that increasing parity, age between 45 years and 49 years, a BMI more than 35, and high levels of anxiety were associated with an increase in urinary incontinence symptoms. Although several studies5–7,13–14 have examined the relationship between urinary incontinence and menopausal stage, very few13 have investigated the relationship between serum E2 change through the menopausal transition and urinary incontinence.
The most important finding of our study was that women who had a sharp decline of E2 levels from premenopause to menopause had a clinically significant decrease in overall incontinence symptom scores. Notably, mean levels of E2 at one point in time were not associated with lower urinary tract symptoms. These findings were surprising because it has been generally accepted that a woman’s urinary symptoms increase with menopause due changes in the pelvic floor from estrogen decline.4,15 Schaffer and Fantl4 demonstrated that a hypoestrogenic state is associated with thinning of the urethral mucosa, changes in urethral closure pressure from loss of urethral sphincter tone, and alteration of the urethrovesical angle. These physiologic changes that occur in urogenital tissue are factors that have been shown to affect the severity of urinary incontinence.
Although these findings were unexpected, they are supported by recent observational studies.5,7 In another large prospective cohort study, Waetjan et al7 reported that women who transitioned from premenopause to menopause had improved overall urinary incontinence. Although not statistically significant, they also reported similar trends in the reduction of stress and urge incontinence. Although we could not assess the change in lower urinary tract symptoms over time because the BFLUTS-SF questionnaire was not administered at each assessment period, we found that women with a sharp decline in E2 levels indicative of a transition from premenopause to menopause had a statistically significant decrease in the severity of urinary incontinence and, more specifically, stress incontinence. We did not, however, see a similar reduction in the severity of urge incontinence. Furthermore, no association was seen between the change in E2 levels through the menopausal transition with voiding and filling urinary symptoms. Slight differences in the results of these studies likely stem from the use of different questionnaires. We should also note that the reduction in stress incontinence score seen in our cohort was below our a priori defined clinical meaningful change in symptom score of one point; therefore, whether this finding is clinically important warrants further investigation.
When adjusting for the effect of the menopausal transition, age was an independent risk factor for incontinence symptoms. In particular, we found women between the ages of 45 years and 49 years had a clinically significant increase in incontinence symptoms. In our study, this age group remained an independent risk factor for worsening urinary incontinence when controlling for E2 hormone change through the menopausal transition. We believe this change in score is clinically significant because a change in level of symptom severity by one response category can have a clinically meaningful effect on a woman’s quality of life. Other studies have supported this finding, because women between the ages of 45 years and 49 years have been found to have the highest prevalence rates of urinary incontinence in comparison with other decades of life.2–3,16 Sampselle et al6 reported a similar increase in the odds of urinary incontinence by 18% for women between the ages of 48 years and 52 years in comparison with younger women in the late reproductive years (aged 42–47 years).
The effect of E2 on target tissues in the bladder and pelvic floor remain unclear. Multiple studies have investigated the effect of exogenous estrogen hormonal therapy with conflicting results. Some studies have demonstrated improvements in lower urinary tract symptoms17–19 with treatment, whereas others have demonstrated worsening urinary symptoms.20–21 A possible explanation for these clinical findings could be that the distribution of estrogen receptors in the bladder and pelvic floor changes during the different menopausal stages.22–24 This change in distribution of E2 receptors could explain the different clinical responses to estrogen therapy seen in women of varying age and menopausal stages. It might also explain our finding that decreased endogenous E2 improves incontinence symptoms. In our cohort of women, we excluded women who were receiving exogenous estrogen supplementation. Hence, we were able to investigate how the natural change in E2 levels through the menopausal transition affect lower urinary tract symptoms. Further research is needed to determine the specific relationship between E2 hormone levels, advancing age, and lower urinary tract symptoms.
We also found that a BMI more than 35, increasing parity, and anxiety were independently associated with an increase in urinary incontinence symptoms. A BMI more than 35 and moderate to high anxiety were also associated with an increase in urinary filling symptoms. These findings are consistent with those reported in previous studies.25–26 Dwyer et al26 reported that obesity was significantly more common in women with stress and urge incontinence. The excess weight in these individuals has been thought to create excess strain on the pelvic floor muscles and nerves, resulting in these urinary symptoms. Anxiety has been previously associated with many types of incontinence (stress, mixed, urge) and lower urinary tract symptoms.25 In this study we found an association between high anxiety and lower urinary tract symptoms. It is not clear whether women with anxiety simply are more likely to report lower urinary tract symptoms, whether lower urinary tract symptoms causes more anxiety, or whether anxiety somehow worsens lower urinary tract symptoms. Due to the small number of women with high anxiety (n=10) in our study cohort, a longitudinal assessment of anxiety and lower urinary tract symptoms would help to further clarify this relationship.
Our study has a number of strengths. This study is unique in that we have the ability to investigate the relationship between the changes in E2 levels over time in premenopausal women as they transition to the different stages of menopause. Furthermore, the focus of our study was to investigate not only urinary incontinence but also the full spectrum of lower urinary tract symptoms such as bladder filling and voiding dysfunction symptoms. None of the prior longitudinal studies have used a validated questionnaire (BFLUTS-SF) that focuses on urinary incontinence and lower urinary tract symptoms. Additionally, 10 years of longitudinal data on E2 hormone level change through the menopausal transition is rare. By investigating E2 change during the menopausal transition, we hope to provide some insight into the relationship between E2 levels and lower urinary tract symptoms. This may help generate further investigations that explore a biologically plausible mechanism for the conflicting results of multiple trials examining the effect of hormonal therapy and urinary incontinence.
Our study also has a number of limitations. We do not have information available on lower urinary tract symptoms from the earlier assessment points because the BFLUTS-SF was added to the latest assessment period of the Penn Ovarian Aging Study. Therefore, we were unable to assess change in lower urinary tract symptoms during the menopausal transition. However, by using a regression coefficient (slope) to describe the trend of change in a woman’s E2 levels through the menopausal transition, we were able to investigate the relationship between E2 change over time and lower urinary tract symptoms. Currently, only 71% of the study participants remained active in the parent Penn Ovarian Aging study at assessment period 11 and therefore were eligible for inclusion into this study. It is possible that this loss to follow-up may introduce selection bias and affect the generalizability of the results. However, a prior comparison of baseline values between active and inactive subjects in this cohort study found no differences in E2 levels and racial and demographic characteristics.27 It is also possible that some of our significant findings may be due to chance, because there were multiple endpoints used in this study design.
Severity of urinary incontinence symptoms has been found to be strongly associated with a decline in E2 levels and ages between 45 years and 49 years. Unlike previously held viewpoints that both age and the menopausal transition worsen incontinence symptoms, we found that a sharp decline in E2, reflective of the progression through the menopausal transition to be protective. Further investigations are needed to elucidate the effect of E2 change, advancing age, and severity of lower urinary tract symptoms.
1. Sommer P, Bauer T, Nielsen KK, Kristensen ES, Hermann GG, Steven K, et al. Voiding patterns and prevalence of incontinence in women. A questionnaire survey. Br J Urol 1990;66:12–5.
2. Yarnell JW, Voyle GJ, Richards CJ, Stephenson TP. The prevalence and severity of urinary incontinence in women. J Epidemiol Community Health 1981;35:71–4.
3. Thomas TM, Plymat KR, Blannin J, Meade TW. Prevalence of urinary incontinence. Br Med J 1980;281:1243–5.
4. Schaffer J, Fantl JA. Urogenital effects of the menopause. Baillieres Clin Obstet Gynaecol 1996;10:401–17.
5. Sherburn M, Guthrie JR, Dudley EC, O’Connell HE, Dennerstein L. Is incontinence associated with menopause? Obstet Gynecol 2001;98:628–33.
6. Sampselle CM, Harlow SD, Skurnick J, Brubaker L, Bondarenko I. Urinary incontinence predictors and life impact in ethnically diverse perimenopausal women. Obstet Gynecol 2002;100:1230–8.
7. Waetjen LE, Feng WY, Ye J, Johnson WO, Greendale GA, Sampselle CM, et al. Factors associated with worsening and improving urinary incontinence across the menopausal transition. Obstet Gynecol 2008;111:667–77.
8. Jackson S, Donovan J, Brookes S, Eckford S, Swithinbank L, Abrams P. The Bristol Female Lower Urinary Tract Symptoms questionnaire: development and psychometric testing. Br J Urol 1996;77:805–12.
9. Brookes ST, Donovan J, Wright M, Jackson S, Abrams P. A scored form of the Bristol Female Lower Urinary Tract Symptoms questionnaire: data from a randomized controlled trial of surgery for women with stress incontinence. Am J Obstet Gynecol 2004;191:73–82.
10. Soules MR, Sherman S, Parrott E, Rebar R, Santoro N, Utian W, et al. Executive summary: Stages of Reproductive Aging Workshop (STRAW). Fertil Steril 2001;76:874–8.
11. Zung WW. A rating instrument for anxiety disorders. Psychosomatics 1971;12:371–9.
12. Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol 1993;138:923–36.
13. Dennerstein L, Dudley EC, Hooper JL, Guthrie JR, Burger HG. A prospective population-based study of menopausal symptoms. Obstet Gynecol 2000;96:351–8.
14. Hannestad YS, Rortviet G, Sandvik H, Hunskaar S. A community-based epidemiological survey of female urinary incontinence: the Norwegian EPINCONT study. Epidemiology of incontinence in the County of Nordtrondelag. J Clin Epidemiol 2000;53:1150–7.
15. Burger HG, Dudley EC, Hopper JL, Shelley JM, Green A, Dennerstein L, et al. The endocrinology of menopausal transition: a cross-sectional study of a population-based sample. J Clin Endocrinol Metab 1995;80:3537–45.
16. Rekers H, Drogendijk AC, Valkenburg H, Riphagen F. Urinary incontinence in women from 35 to 79 years of age: prevalence and consequences. Eur J Obstet Gynecol Reprod Biol 1992;43:229–34.
17. Moehrer B, Hextall A, Jackson S. Oestrogens for urinary incontinence in women. The Cochrane Database of Systematic Reviews 2003, Issue 2. Art. No.: CD001405. DOI: 10.1002/14651858.CD001405.
18. Andersson KE, Appell R, Cardozo LD, Chapple C, Drutz HP, Finkbeiner AE, et al. The pharmacological treatment of urinary incontinence. BJU Int 1999;84:923–47.
19. Sarkar PK, Ritch AE. Management of urinary incontinence. J Clin Pharm Ther 2000;25:251–63.
20. Grady D, Brown JS, Vittinghoff E, Applegate W, Varner E, Snyder T, et al. Postmenopausal hormones and incontinence: the Heart and Estrogen/Progestin Replacement Study. Obstet Gynecol 2001;97:116–20.
21. Hendrix SL, Cochrane BB, Nygaard IE, Handa VL, Barnabei VM, Iglesia C, et al. Effects of estrogen with or without progestin on urinary incontinence. JAMA 2005;293:935–48.
22. Schreiter F, Fuchs P, Stockamp K. Estrogenic sensitivity of alpha-receptors in the urethra musculature. Urol Int 1976;31:13–9.
23. Gebhart JB, Rickard DJ, Barrett TJ, Lesnick TG, Webb MJ, Podratz KC, et al. Expression of estrogen receptor isoforms alpha and beta messenger RNA in vaginal tissue of premenopausal and postmenopausal women. Am J Obstet Gynecol 2001;185:1325–30.
24. Bergman A, Karram MM, Bhatia NN. Changes in urethral cytology following estrogen administration. Gynecol Obstet Invest 1990;29:211–3.
25. Lim JR, Bak CW, Lee JB. Comparison of anxiety between patients with mixed incontinence and those with stress urinary incontinence. Scand J Urol Nephrol 2007;41:403–6.
26. Dwyer PL, Lee ET, Hay DM. Obesity and urinary incontinence in women. Br J Obstet Gynaecol 1988;95:91–6.
27. Nelson DB, Sammel MD, Freeman EW, Liu L, Langan E, Gracia CR. Predicting participation in prospective studies of ovarian aging. Menopause 2004;11:543–8.