Click on the links below to access all the Data Supplements for this article.
Please note that Data Supplement files may launch a viewer application outside of your web browser.
Pelvic organ prolapse is common, such that one in nine American women will undergo surgical repair for pelvic organ prolapse and/or urinary incontinence (UI) in her lifetime.1 Although some women undergo surgery to alleviate a bulge, others choose surgery because of bothersome pelvic floor symptoms. In cross-sectional studies, symptoms are common in women with early (stages I and II) prolapse but often do not correlate well with anatomic levels of prolapse.2–4 However, such studies cannot inform us about whether changes in vaginal support are associated with the presence of symptoms.
We previously reported that only obstructive urinary and vaginal bulge symptoms were associated with the level of vaginal descent in a cross-sectional analysis of older women.4 This study's objective was to determine whether vaginal descent progression was associated with pelvic floor symptoms in the same women when followed over time.
MATERIALS AND METHODS
This longitudinal study included women enrolled in the Women's Health Initiative (WHI) Hormone Therapy randomized clinical trial at one Midwestern site. All women were postmenopausal and had a uterus. As previously described,4,5 participants completed a questionnaire and underwent a pelvic examination with pelvic organ prolapse quantification (POP-Q)6 at yearly research visits over a 4-year period. In year 4, data could not be collected in many women because the overall WHI trial was halted prematurely.7 One woman underwent surgery for prolapse (hysterectomy and rectocele repair) during the course of the study and was excluded. All remaining women who completed a questionnaire and examination in at least 2 of the 4 years were included in this analysis. This study was approved by the University of Iowa Institutional Review Board, and each subject completed an informed consent document.
As previously reported,8 we modified 30 items from the Pelvic Floor Distress Inventory, a validated quality-of-life instrument that is condition specific for women with pelvic floor disorders.9 Because our main purpose was to determine whether specific symptoms were present or absent in our group of non–care-seeking women, we modified the Pelvic Floor Distress Inventory stem so that each item functioned as a yes/no symptom-screening question. We also constructed symptom groups, similar to the subscale groupings described by Barber et al,9 as follows: overactive bladder, bladder pain, obstructive bladder, prolapse, obstructive bowel, bowel incontinence, and bowel pain. We calculated simple additive scores for each symptom group by assigning a score of 0 to the absence and 1 to the presence of each symptom within a group. Symptom group scores were then categorized into two or three levels to more evenly distribute women between levels.
Pelvic organ prolapse quantification measurements and staging were performed as recommended by the International Continence Society.6 As previously detailed, the POP-Q examinations were performed by one of two experienced urogynecology research nurses while the subject was in dorsal lithotomy position.10 All measurements, except total vaginal length, were taken during maximal Valsalva maneuver. Measurements were recorded as whole numbers; changes in vaginal descent were thus assessed in 1-cm increments. Women were not told the results of the POP-Q examination. The women also provided information about obstetric, medical, and surgical histories in the written questionnaire. Other demographic and background data were abstracted from the WHI database.
Continuous variables were summarized by mean and standard deviation or by median and range for variables with nonnormal distributions (as determined by the Shapiro-Wilk test). Multivariable models were created to study associations between vaginal descent (independent variable) and individual symptoms or symptom group scores (dependent variable), controlling for other risk factors found to be associated with symptoms in our prior work.8 Longitudinal data analysis based on generalized linear (logistic) models was used, and parameters were estimated via the generalized estimating equation,11 as prolapse measures and symptom outcomes were repeatedly measured over time. For symptom scores categorized in three levels, a dichotomous outcome of the lowest compared with the highest level was used.
We a priori planned to create models for four individual symptoms and for all symptom group scores. We included the individual symptoms of “a sensation of vaginal bulging” and “see or feel a vaginal bulge,” because these symptoms were most strongly associated with prolapse measures in previous cross-sectional analyses.4,12 We also specifically studied individual symptoms of stress urinary incontinence and urge urinary incontinence (both defined as at least weekly incontinence), due to their high clinical relevance.
For each dependent outcome (four individual symptoms and seven symptom group scores), we created two separate models: one model that included the point of maximal vaginal descent as the independent prolapse variable and a second model that included either point Ba, Bp, or C as the independent prolapse variable. For symptoms or scores related to the bladder, we included point Ba as a measure of anterior vaginal descent. For symptoms or scores related to colorectal function, we included point Bp as a measure of posterior vaginal descent. Lastly, for the individual vaginal bulge symptoms and the prolapse symptom score, we included point C as a measure of apical prolapse.
Vaginal descent variables were initially included as continuous variables. Associations between the vaginal descent variable and the symptom outcome were then studied graphically, and vaginal descent variables were dichotomized when the associations appeared nonlinear.
Potential risk factors included in the models were categorized as following: age and body mass index (BMI) as continuous variables, and exercise (at least weekly), coffee drinking, and current smoking as dichotomous variables. Covariates added to the initial models included time, BMI, and age (all models), exercise and coffee drinking (in the bladder symptom models), and smoking (in the prolapse symptom models). Results did not change significantly when the latter three covariates were dropped from the models, so for the sake of consistency, the final models all controlled for time, BMI, and age.
Statistical analyses were performed with SAS 9.1 (SAS Institute, Inc., Cary, NC). Associations were considered significant at a P<.05 level. Our sample size was predetermined by the number of women in this ancillary study.
Two hundred seventy (80%) of 337 eligible women were enrolled and examined in year 1. Of those, 260 women completed at least two questionnaires with corresponding examinations and constituted the population for this longitudinal analysis. In years 1, 2, 3, and 4, 260, 259, 249, and 208 women completed questionnaires, and 260, 242, 212, and 86 women completed examinations, respectively. Baseline participant characteristics are presented in Table 1. Other characteristics of the study sample have been previously described.5 Five women (1.9%) reported prior surgery for pelvic organ prolapse on enrollment (occurring from 3 to 31 years earlier). Four women reported pessary use for pelvic organ prolapse or UI at one or more visits during the study. The majority of the women had POP-Q stages I and II prolapse. As previously reported, when prolapse was defined as vaginal descent reaching the hymen or beyond, the 1-year and 3-year prolapse incidences were 26% and 40%.10
In year 1, 53 (20.4%) women reported weekly stress urinary incontinence, and 53 (20.4%) reported weekly urge urinary incontinence. Thirty-six (13.9%) women reported both types of incontinence. Twelve (4.6%) reported a “sensation of a bulge,” and 10 (3.9%) reported “seeing/feeling a vaginal bulge.” Symptom group score results at baseline and the definitions used in categorizing each are shown in Table 2. The overactive bladder symptoms were most frequent; this symptom group was the only one in which more than 50% of the participants reported at least one symptom (that is, the median group score greater than 0) in every year.
Table 3 summarizes the associations between vaginal descent and symptoms found in the longitudinal models. Increasing maximal vaginal descent was strongly associated with the “see/feel a bulge” symptom and also associated with “sensation of protrusion or bulging.” A weaker association was found between increasing apical descent (C) and “see/feel a bulge.” Increasing vaginal descent (both anterior [Ba] and maximal) was associated with bladder pain and obstructive bladder symptoms. Maximal vaginal descent (but not apical descent) was positively associated with the general prolapse score. Weaker positive associations were seen between vaginal descent measures and bowel symptoms, including maximal vaginal descent with the obstructive bowel score and posterior vaginal descent (Bp) with the bowel incontinence score. Vaginal descent was not associated with stress urinary incontinence, urge urinary incontinence, the overactive bladder score, or the bowel pain score (data not shown).
In some cases, the vaginal descent measures were not associated linearly with symptom outcomes. Such nonlinear associations included those of the maximal vaginal descent with “see/feel a bulge” and with “sensation of protrusion or bulging,” maximal and anterior vaginal descent with the bladder pain symptom score, and maximal descent with the prolapse symptom score. In all of these cases, the linearity plot suggested a “threshold” effect when vaginal descent approached 0 cm (the hymen). Therefore, the vaginal descent variable was dichotomized in these cases (Table 3). No women with maximal vaginal descent less than 0 cm reported the symptom of “see/feel a bulge,” so in order for the model to converge, we had to categorize this variable as less than +1 cm or +1 cm or greater (rather than 0 cm).
In addition to vaginal descent, several other risk factors (including age, BMI, coffee drinking, and time) were found to be independently associated with symptoms in the longitudinal models. These results are presented in Table 4. Because odds ratios were similar across models with the same symptom outcome (dependent variable), only those from the models with the greater number of covariates and that included maximal vaginal descent are listed.
In contrast to our previous cross-sectional findings, this longitudinal analysis suggests that increases in vaginal descent are positively associated with several types of pelvic floor symptoms in postmenopausal women with stages I–II pelvic organ prolapse.
We and others have suggested that vaginal descent to the hymeneal ring is “normal” (that is stages I and early II) because 1) descent to this level is common, occurring in more than half of older, parous women, 2) most women with descent at this level do not progress to stages III and IV prolapse, and 3) descent at this level correlates poorly with the presence of symptoms.4,10,12,13 However, in this longitudinal analysis, minor increases in vaginal descent were associated with an array of pelvic floor symptoms, despite the fact that women largely still remained in this “normal” anatomic range. It is possible that women adjust to symptoms over time, such that, in cross-sectional studies, they no longer report symptoms as present unless they progress to a higher level of severity. Or perhaps, by including repeated measurements over time, this analysis effectively increased our power to detect weaker associations between vaginal descent and symptoms that were not detected in previous studies, in addition to the stronger ones seen in previous work.
Other aspects of our results are similar to past findings in cross-sectional studies. Vaginal descent progression was most strongly associated with vaginal bulge symptoms, especially “see/feel a vaginal bulge.” This single symptom has been demonstrated in several previous studies to be most predictive of pelvic organ prolapse, although its sensitivity has varied based on the prevalence of prolapse in the studied population.12–16 We also found a “threshold” effect in the association of vaginal descent with the vaginal bulge symptoms and with the prolapse symptom group score. Similar to past reports, vaginal descent reaching the hymen or beyond was strongly associated with bulge symptoms, compared with women with vaginal descent that did not reach the hymen.12–14 A similar effect was seen in this study between vaginal descent and the bladder pain score, but not for other symptoms, including the obstructive bladder and bowel symptom scores.
The positive associations seen between vaginal descent and most bladder and bowel symptoms were weaker than that seen for the vaginal bulge symptoms (odds ratios less than 2), suggesting stronger relationships between other factors and these symptoms. For instance, the increased risk for obstructive bladder symptoms associated with a 1-cm increase in maximal vaginal descent (odds ratio [OR] 1.6) was similar to the risk associated with a 5-year increase in age (OR 1.8) and less than that associated with coffee drinking (OR 4.0).
Strengths of this study include its prospective design, use of a validated instrument to assess prolapse, and the longitudinal analysis, which is unique in the study of pelvic floor symptoms and prolapse. We acknowledge limitations to our results as well. Our study population consists largely of white, older, and parous women, which limits generalizability. However, this is the population most likely to undergo surgery for prolapse. Additionally, although we found that symptoms were associated with progression of vaginal descent, we did not measure the severity or level of bother associated with reported symptoms. A previous study found that, at least for the “see/feel a vaginal bulge” symptom, considering the degree of bother in addition to the presence or absence of the symptom did not improve the discriminative ability of the symptom, reassuring us of the validity of our results.13
Finally, the results from this longitudinal study do not allow us to draw conclusions about the utility of treating early stages of pelvic organ prolapse in terms of symptom resolution. However, for those women seeking treatment for symptoms and early prolapse, it seems likely that surgery would not alleviate symptoms in many cases, given the association between symptoms and non–prolapse-related factors (such as coffee drinking or obesity). On the other hand, given the low risks, it may be reasonable to suggest that, when new symptoms arise coincident with increased vaginal descent, initiating a pelvic muscle strengthening program may help to alleviate both prolapse and symptoms.17 Our findings, which suggest a stronger link between symptoms and vaginal descent thanpreviously described, emphasize the need for future research on conservative therapies for “early” pelvic organ prolapse in terms of symptom outcomes and, ultimately, the prevention of vaginal descent progression.
1. Olsen AL, Smith VJ, Bergstrom JO, Colling JC, Clark AL. Epidemiology of surgically managed pelvic organ prolapse and urinary incontinence. Obstet Gynecol 1997;89:501–6.
2. Ellerkmann RM, Cundiff GW, Melick CF, Nihira MA, Leffler K, Bent AE. Correlation of symptoms with location and severity of pelvic organ prolapse. Am J Obstet Gynecol 2001;185:1332–7; discussion 1337–8.
3. Burrows LJ, Meyn LA, Walters MD, Weber AM. Pelvic symptoms in women with pelvic organ prolapse. Obstet Gynecol 2004;104:982–8.
4. Bradley CS, Nygaard IE. Vaginal wall descensus and pelvic floor symptoms in older women. Obstet Gynecol 2005;106:759–66.
5. Nygaard I, Bradley C, Brandt D. Pelvic organ prolapse in older women: prevalence and risk factors. Obstet Gynecol 2004;104:489–97.
6. Bump RC, Mattiasson A, Bo K, Brubaker LP, DeLancey JO, Klarskov P, et al. The standardization of terminology of female pelvic organ prolapse and pelvic floor dysfunction. Am J Obstet Gynecol 1996;175:10–7.
7. 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.
8. Bradley CS, Kennedy CM, Nygaard IE. Pelvic floor symptoms and lifestyle factors in older women. J Womens Health (Larchmt) 2005;14:128–36.
9. Barber MD, Kuchibhatla MN, Pieper CF, Bump RC. Psychometric evaluation of 2 comprehensive condition-specific quality of life instruments for women with pelvic floor disorders. Am J Obstet Gynecol 2001;185:1388–95.
10. Bradley CS, Zimmerman MB, Qi Y, Nygaard IE. Natural history of pelvic organ prolapse in postmenopausal women. Obstet Gynecol 2007;109:848–54.
11. Diggle P, Liang K, Zeger S. Analysis of longitudinal data. New York (NY): Oxford University Press; 1994. p. 143–5.
12. Swift S, Woodman P, O'Boyle A, Kahn M, Valley M, Bland D, et al. Pelvic Organ Support Study (POSST): the distribution, clinical definition, and epidemiologic condition of pelvic organ support defects. Am J Obstet Gynecol 2005;192:795–806.
13. Barber MD. Symptoms and outcome measures of pelvic organ prolapse. Clin Obstet Gynecol 2005;48:648–61.
14. Tan JS, Lukacz ES, Menefee SA, Powell CR, Nager CW. Predictive value of prolapse symptoms: a large database study. Int Urogynecol J Pelvic Floor Dysfunct 2005;16:203–9.
15. Lukacz ES, Lawrence JM, Buckwalter JG, Burchette RJ, Nager CW, Luber KM. Epidemiology of prolapse and incontinence questionnaire: validation of a new epidemiologic survey. Int Urogynecol J Pelvic Floor Dysfunct 2005;16:272–84.
16. Barber MD, Neubauer NL, Klein-Olarte V. Can we screen for pelvic organ prolapse without a physical examination in epidemiologic studies? Am J Obstet Gynecol 2006;195:942–8.
17. Piya-Anant M, Therasakvichya S, Leelaphatanadit C, Techatrisak K. Integrated health research program for the Thai elderly: prevalence of genital prolapse and effectiveness of pelvic floor exercise to prevent worsening of genital prolapse in elderly women. J Med Assoc Thai 2003;86: 509–15.