The levator ani muscles are unique striated muscles which play a critical role in pelvic organ support. Their activity automatically adjusts to variations in posture and abdominal pressure1,2 to provide upward support to the pelvic viscera. The action of the levator minimizes the load on connective tissues that attach the organs to the pelvis. Failure of these supportive structures leads to pelvic organ prolapse and surgery for over 200,000 American women each year.3 The identification of discrete injuries to the muscles and connective tissues that lead to pelvic organ prolapse is needed to improve our understanding of the pathophysiology of prolapse and to develop strategies of prevention.
Vaginal birth confers a four- to 11-fold increase in risk for developing prolapse among parous women4 and is the single most important modifiable preventive factor for this condition.5 Recent studies have demonstrated that levator ani defects occur after vaginal birth.6–8 It is not appreciated whether these defects lead to pelvic organ prolapse later in life. Potential for pelvic floor injury during vaginal birth is a major factor driving the current debate for prelabor Cesarean delivery on maternal request9 recently highlighted in a National Institutes of Health State-of-the-Science conference (http://consensus.nih.gov). The strategy of prelabor Cesarean delivery, however, could subject nine women to an operation for every woman that might benefit later in life.5 Efforts to identify women at high risk are hindered by a lack of understanding regarding how vaginal birth or specific determinants such as levator ani injury lead to prolapse. If the type and mechanism of birth-induced injury could be identified, prevention strategies could be developed to target high-risk women and to avoid subjecting lower-risk women to unnecessary intervention. The present study seeks to estimate whether the rate of levator ani defects among women with prolapse is higher than that observed among controls and to quantify the effect of such a defect on pelvic floor muscle function.
PARTICIPANTS AND METHODS
Between November 2000 and October 2004, 151 women with prolapse and 135 controls were recruited to participate in a University of Michigan institutional review board–approved, case–control study with group matching of pelvic organ prolapse. Cases were recruited from the Urogynecology Clinic at our institution and controls by advertisement in the local communities. These groups were recruited to be of similar age, parity, self-reported race, and hysterectomy status. A pelvic organ prolapse quantification (POP-Q)10 examination was performed and used to determine eligibility for inclusion in the case and control groups. To be included as a case, prolapse of one of the following—a vaginal wall, the hysterectomy scar, or the cervix—had to protrude at least 1 cm above the hymen during Valsalva. The inclusion of individuals in the control group with apical support as described is in the spirit of the precedent established by an National Institutes of Health convened workgroup in 2001.11 Only two individuals in the control group had apical support just 1cm above the hymen.
Women were excluded from the case group if a vaginal wall, the hysterectomy scar, or the cervix was not at least a centimeter below the hymen. Women were excluded from the control group if any POP-Q point was greater than minus 1 cm or if they reported symptoms of or demonstrated stress urinary incontinence during testing. Women who had undergone hysterectomy 1 year before enrollment, who had any surgery for urinary incontinence or prolapse, who had genital anomalies, or who had a history of radiation therapy were excluded, because these factors might distort the pelvic organ support tissues. They were also excluded if they had diseases (eg, sickle cell anemia) that might increase the risks of infection from testing.
Subjects completed a questionnaire regarding medical, surgical, reproductive, and social and family histories. They were asked to indicate whether a mother, a grandmother, or a sister was affected by pelvic organ prolapse. A questionnaire regarding symptoms of pelvic floor dysfunction specifically included the following two items regarding pelvic organ prolapse: 1) Do you have a sensation of bulging or protrusion from the vaginal area? 2) Do you have a bulge of something falling out that you can see or feel in the vaginal area?
Subjects were asked to identify whether they had such symptoms “never,” “on an occasional day,” “on most days,” or “on every day.” For the purpose of illustration, the answers were dichotomized between “never” and “on an occasional day.”
Subjects' height and weight were recorded. A pelvic examination was performed with the subject in a semirecumbent lithotomy position. Pelvic organ prolapse quantification7 (POP-Q) was used to quantify vaginal support. The subject was instructed in pelvic floor muscle contraction until satisfactory performance of a contraction was demonstrable as previously described.1 Levator ani muscle function was assessed with a vaginal speculum specially designed to record the isometric forces acting on the anterior and posterior bills of the speculum at rest and with pelvic floor contraction. Recordings of vaginal closure force at rest represented the passive closure forces from connective tissue and resting muscle tone. Women were asked to perform a maximal pelvic muscle contraction and the amount of force above resting tone was recorded as the augmentation of vaginal closure force.
Each woman underwent magnetic resonance imaging (MRI) in the axial, sagittal, and coronal planes using a fast spin proton density technique (echo time: 15 ms, repetition time: 4,000 ms) in the supine position. Scans were performed on a 1.5 T superconducting magnet (Signa; General Electric Medical Systems, Milwaukee, WI). Slice thickness was 4 mm, with a gap of 1 mm, yielding 5-mm image spacing. A 160×160 mm field of view and 256×256 imaging matrix were used.
Levator muscle defects were graded independently on MR scans by two examiners blinded to subject status using a system previously described for evaluating birth-associated damage.6 The left and right muscles were scored separately. An example of different degrees of muscle defect is shown in Figure 1. A score of “0” was assigned if no damage was visible, “1” if less than half of the muscle was missing, “2” if more than half, and “3” if the complete muscle bulk was lost. When reviewers differed on scores, scans were reviewed together to assign a final score. The total score was the sum of the two sides, ranged from 0 to 6, and was categorized as follows: 0 normal or no defect; 1–3 minor defect; 4–6 major defect. Women with a unilateral 3 were considered major defects. The interrater reliability of this scoring system was recently described.12
An appropriate sample size was estimated as follows: In a previous study, we observed that 17% of primiparous women had a levator ani defect.7 Anticipating that the odds ratio for a levator ani defect in prolapse cases compared with controls would be at least 2.0, we proposed a sample size of 150 women per group, with the probability of a defect being, respectively, 0.17 and 0.31 for cases and controls. The 2 by 2 contingency table implied by these assumptions gave us an odds ratio of 2.1 with a power of 0.81 to detect a significant difference with α=0.05.
Means, standard deviations, and frequency distributions for discrete variables were used to characterize the groups. Case–control comparisons for continuous outcomes were made using general linear models which included age, body mass index (BMI), hormone replacement status, race, family history of prolapse, and vaginal parity as covariates. Case–control comparisons for discrete outcomes were made using logistic regression models with the covariate controls. Other continuous outcome measures were analyzed using general linear models that included case–control status, levator ani defect status, and relevant covariates.
Table 1 indicates that the groups were similar in age, race, hysterectomy status, menopausal status, and BMI. Despite attempts to match for parity, there was a small (less than one birth) difference in the number of vaginal deliveries due to difficulties in identifying women of higher parity without any degree of prolapse. Women with prolapse were also significantly more likely to have had forceps delivery, report a family history of prolapse, and to be taking hormonal therapy.
Pelvic organ prolapse quantification stages of cases with prolapse were: stage II, 46 (30.5%), stage III, 101 (66.9%), and stage IV, 4 (2.6%). Because this staging did not provide satisfactory sized groups, for the purposes of this report, we created three more equally distributed groups of cases with prolapse: Small (+1 cm, n=46), medium (+2 to +3 cm, n=64) and large (+4 cm or more, n=41). The most dependent point, or leading edge, was the anterior compartment in 62.3%, the apical compartment in 19.9%, and the posterior compartment in 17.9%.
In our data, we found a strong relationship between prolapse and the occurrence of major levator ani defects, with a major defect occurring in 55% of cases and 16% of controls (Fig. 2). Women with prolapse had roughly similar rates of minor defects (16% compared with 22%), and substantially fewer normal-appearing muscles (29% compared with 62%). When we modeled the occurrence of a major levator ani defect in the context of a multivariable logistic regression as a function of prolapse status and relevant covariates (ie, age, BMI, hormone therapy status, race, family history of prolapse, and vaginal parity), we found prolapse status was significant (P<.001) and was associated with an adjusted odds ratio of 7.3 (95% confidence interval 3.9–13.6).
We also found a strong relationship between a self-reported history of having had a forceps delivery and the occurrence of a major levator ani defect. Forty-four percent of prolapse cases in our sample reported having a forceps deliveries in comparison with 23% for control cases (P<.001). Furthermore, 53% of all women delivered by forceps had a major levator ani defect in comparison with 28% for women not reporting forceps delivery. When we modeled the occurrence of a major levator ani defect in the context of a logistic regression including forceps status and relevant covariates, we found that women reporting delivery by forceps (P<.001) had an odds ratio of 3.4 (95% confidence interval 1.95–5.78) for having a levator defect.
In the sample of women with prolapse, the most striking difference among the defect status groups shown in Figure 3 is the difference in the proportion of small prolapses (ie, 52.3% for cases without any defect and, respectively, 20.8% and 21.7% for minor and major defect cases). The χ2 test for the implied 2 by 3 contingency was significant, P=.001. Seventy-five percent of women with an anterior prolapse and 80% of those with an apical prolapse had minor or major levator ani defects, whereas only 48% of those with a posterior prolapse had such defects (P=.014).
The genital hiatus was 50% longer in women with prolapse compared with controls (Table 1). The difference in mean genital hiatus (±standard deviation) between cases and controls was present at each of the three levels of muscle defect: major defects (4.9±1.3 cm compared with 3.3±1.0 cm, P<.001), minor defects (4.6±1.3 cm compared with 3.2±1.1 cm, P<.001) and no defects (4.4±1.3 cm compared with 3.0±1.0 cm, P<.001).
Figure 4, shows that the vaginal closure force in the six groups defined by levator ani defect and prolapse status had similar resting closure force. The mean increase above resting during a maximal contraction was higher in women with no defects than in women with minor or major defects in both the prolapse and control groups. Mean maximal contraction was lower for women with prolapse compared with controls for both women with defects and women without defects. Mean maximal contraction was similar for major and minor defect cases in both the prolapse and nonprolapse groups.
Given these observations, we modeled maximal contraction using a general linear model in which the primary variables of interest were case–control status and levator ani defect status (ie, no defect compared with any defect) in addition to covariate controls. Comparing all women with prolapse to controls, this analysis showed that women with prolapse had a lower adjusted maximal contraction (2.0 N) than controls (3.2 N, P<.001). Similarly, women with a defect had a lower adjusted maximal contraction (2.0 N) than women without a defect (3.1 N), P<.001.
In addition to these overall tests, we carried out post hoc Scheffé adjusted tests of pair-wise comparisons involving the four adjusted means defined by prolapse status and defect status where minor and major are combined into a defect group. Among cases, those with no defects had higher adjusted mean maximal contraction force than women with defects (2.5 N compared with 1.5, P=.014). Among controls, women with no defect had an adjusted higher force than women with a defect (3.8 N compared with 2.6 N, P=.003). Considering women with no defect, cases had a lower than controls (2.5 N compared with 3.8 N, P=.004) while in women with a levator ani defect, cases had a lower adjusted mean maximal contraction force than controls (1.5 N compared with 2.6 N, P=.006).
The proportion of cases with major levator ani muscle defects was significantly higher than that observed among age-, race-, and hysterectomy-matched controls. More than one half of the women with prolapse had major defects. These findings in a large case–control study with group matching establish that levator impairment is associated with prolapse. They also suggest that the reduced muscle thickness found in an earlier study of women with and without prolapse13 was not due to a biased comparison of older women with prolapse and younger control group.14 Furthermore, this study quantifies how vaginal closure force is affected by the presence of pelvic organ prolapse and muscle defect status. Women with prolapse were unable to augment vaginal closure force as well as controls, and women with muscle defects were also unable to augment vaginal closure force as well as those with normal muscles in the case and control cohorts.
The association of levator ani injury with prolapse is a plausible factor in the pathophysiology of prolapse but does not account for all pelvic organ prolapse. The fact that 30% of the women with prolapse had no evidence of a muscle defect on MRI supports the fact that the disease process involves other factors as well. This fact does not diminish the importance of levator ani injury but is a reminder that multiple aspects of the system underlie failure. For instance, it is plausible that failure in one element (levator ani muscle) could lead to increased demands on other components (connective tissue and smooth muscle) which may eventually fail as well. The findings of the present study establish that the injury to the levator ani plays a significant role in prolapse and add to a growing body of data concerning connective tissue abnormalities15,16 and smooth muscle17,18 in the development of prolapse. Future studies will be needed to provide a more complete picture of the interaction between these components in providing pelvic organ support.
The levator ani defects observed in this study of middle-aged women were similar to those seen after vaginal birth in a cohort of younger patients.6,7 The recruitment of this older cohort was necessary to establish an association between levator injury and prolapse because of the latency between birth and the development of the condition. If the findings of these cohorts are taken together, they suggest that there is a relationship between birth-induced injury and the development of pelvic organ prolapse. The large difference between cases and controls in the proportion of levator ani injuries is an indication of their significance and why they are worthy of further study. Future studies will be needed to improve our understanding of how biologic determinants such as levator ani injury can lead to prolapse.
The association of levator ani muscle defects with prolapse and impaired pelvic floor muscle function is consistent with the findings of previous studies concerning levator ani muscle structure and electrophysiology.19–24 Magnetic resonance scans have recently been used to assess the “levator sling muscle gap” (ie, the distance between the pubic bone and remaining muscle) and differences were found between women with pelvic organ prolapse and controls.13,25 Our measurement strategy of assessing the proportion of muscle seen to be missing differs from their observations of the distance between the pubis and remaining muscle, but both studies document decreased muscle volume in the same region.
This study also provides preliminary evidence of exposures that might confer an increased risk for prolapse. Cases were delivered by forceps at a rate twice that of controls. This finding is additional evidence of a relationship between difficult vaginal birth and the development of prolapse.7 Possible mechanisms of injury include excessive stretching of muscle26 and compression or excessive stretching of nerve.27–29 We do not believe that this should necessarily lead to elective cesarean delivery5 but rather to better identification of women at risk, because many factors need to be balanced in management of labor. Future research will be needed to identify the relative contributions of each of these types of damage in causing prolapse.
Another exposure that may lead to higher rates of prolapse is the use of hormone therapy. The proportion of cases using hormone therapy was significantly higher than controls. It is possible that the women with prolapse have hormone therapy prescribed in an attempt to help with the prolapse. However, our data are not consistent with this hypothesis. Case and control cohorts did not differ between the length of time they used hormone therapy. More than two thirds of each cohort reported using it for more than five years. There is some evidence from a small cohort of patients that certain estrogen therapies can have an adverse effect on vaginal support.30 The finding we report may be additional preliminary evidence of such an effect but should be interpreted with caution. This study was not designed to address whether hormone therapy plays a role in the cause of prolapse, and we did not collect information regarding type of hormone therapy used. Finally, a genetic predisposition for prolapse may be present. Cases with prolapse were significantly more likely to report a family history of prolapse in a mother, grandmother, or sister. This could be the result of recall bias among cases, but it may also be a reflection of subtle heritable connective tissue and muscle abnormalities.
There are limitations regarding inferences to be made about this sample of women. Information regarding exposures such as use of forceps during vaginal deliveries, use of hormone therapy, and family history of prolapse are retrospectively collected and this can introduce recall bias, because cases may be more likely than controls to remember such details. A case–control study design does not allow an estimation of the prevalence of levator ani damage among all women because the selection bias is likely to occur with subject recruitment. In this study, women were excluded from the control group if the vagina or cervix was within a centimeter of the hymen, and the sample is therefore not a reflection of the general population. However, we felt that this approach was necessary to avoid differences of opinion regarding cohort assignment. Another advantage of this study design was that it also allowed us to match subjects for confounders in estimating the association between levator injury and prolapse. The groups were matched for age, race, and hysterectomy status but could not be matched exactly for parity because of difficulty in finding controls of high parity meeting the vaginal support inclusion criteria. Having found that levator injuries were present in a significantly higher proportion of cases, we eventually decided that the effort to continue to recruit controls with high parity was not necessary. To account for the difference in parity, logistic regression taking covariates into account was performed. It showed no significant changes in the inferences we made on the total sample and gave us substantial confidence that our results are not due to differences in parity.
Currently, available assessments of pelvic floor characteristics have limitations. The measurement of levator ani function is a somewhat artificial action and may not reflect muscle activity during the great variety of stresses placed on the pelvic floor during an active woman's life. As such, it may not capture all problems in levator function. Differences in vaginal closure force observed among cases and controls when stratified for defect status suggest that other mechanisms, not yet accounted for, are also responsible for some of the difference. Studies using other techniques, now that the location and types of injury present are known, may provide insight about the characteristics of the soft tissue injury associated with prolapse.
The scientific study of pelvic floor dysfunction is in its infancy. Other major discoveries remain to be made. The high prevalence of prolapse,3 frequent recurrence after surgery,31 and the great impact of vaginal birth4 all indicate the need for further research into the cause and prevention of this distressing condition. The findings of this study, by objectively demonstrating levator ani muscle injury and reduced muscle strength in women with prolapse, are important contributions to this field. Future research is needed to both prevent and treat these injuries to reduce the rate of pelvic organ prolapse later in life.
1. Morgan DM, Kaur G, Hsu Y, Fenner D, Guire K, Miller J, et al. Does vaginal closure force differ in the supine and standing positions? Am J Obstet Gynecol 2005;192:1722–8.
2. Shafik A, Doss S, Asaad S. Etiology of the resting myoelectric activity of the levator ani muscle: physioanatomic study with a new theory. World J Surg 2003;27:309–14.
3. Boyles SH, Weber AM, Meyn L. Procedures for pelvic organ prolapse in the United States, 1979-1997. Am J Obstet Gynecol 2003;188:108–15.
4. Mant J, Painter R, Vessey M. Epidemiology of genital prolapse: observations from the Oxford Family Planning Association study. Br J Obstet Gynaecol 1997;104:579–85.
5. Patel DA, Xu X, Thomason AD, Ransom SB, Ivy JS, DeLancey JO. Childbirth and pelvic floor dysfunction: an epidemiologic approach to assessing prevention opportunities at delivery. Am J Obstet Gynecol 2006;195:23–8.
6. DeLancey JO, Kearney R, Chou Q, Speights S, Binno S, et al. The appearance of levator ani muscle abnormalities in magnetic resonance images after vaginal delivery. Obstet Gynecol 2003;101:46–53.
7. Kearney R, Miller JM, Ashton-Miller JA, DeLancey JO. Obstetric factors associated with levator ani muscle injury after vaginal birth. Obstet Gynecol 2006;107:144–9.
8. Dietz HP, Lanzarone V. Levator trauma after vaginal delivery. Obstet Gynecol 2005;106:707–12.
9. Al-Mufti R, McCarthy A, Fisk NM. Obstetricians' personal choice and mode of delivery. Lancet. 1996;347:544.
10. 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.
11. Weber AM, Abrams P, Brubaker L, Cundiff G, Davis G, Dmochowski RR, et al. The standardization of terminology for researchers in female pelvic floor disorders. Int Urogynecol J Pelvic Floor Dysfunct 2001;12:178–86.
12. Morgan DM, Umek W, Stein T, Hsu Y, Guire K, DeLancey JO. Interrater reliability of assessing levator ani muscle defects with magnetic resonance images. Int Urogynecol J 2006; Oct 17 [Epub ahead of print].
13. Hoyte L, Jakab M, Warfield SK, Shott S, Flesh G, Fielding JR. Levator ani thickness variations in symptomatic and asymptomatic women using magnetic resonance-based 3-dimensional color mapping. Am J Obstet Gynecol 2004;191:856–61.
14. Lexell J. Human aging, muscle mass, and fiber type composition. J Gerontol A Biol Sci Med Sci 1995;50:11–6.
15. Jackson SR, Avery NC, Tarlton JF, Eckford SD, Abrams P, Bailey AJ. Changes in metabolism of collagen in genitourinary prolapse. The Lancet 1996;347:1658–61.
16. Chen BH, Wen Y, Li H, Polan ML. Collagen metabolism and turnover in women with stress urinary incontinence and pelvic prolapse. Int Urogynecol J Pelvic Floor Dysfunct 2002;13:80–7.
17. Boreham MK, Wai CY, Miller RT, Schaffer JI, Word RA. Morphometric properties of the posterior vaginal wall in women with pelvic organ prolapse. Am J Obstet Gynecol 2002;187:1501–8.
18. Boreham MK, Wai CY, Miller RT, Schaffer JI, Word RA. Morphometric analysis of smooth muscle in the anterior vaginal wall of women with pelvic organ prolapse. Am J Obstet Gynecol 2002;187:56–63.
19. Halban J, Tandler J. Anatomie und Aetiologie der Genitalprolapse beim Weibe. Vienna (Austria): Wilhelm Braumueller; 1907.
20. Berglas B, Rubin IC. Study of the supportive structures of the uterus by levator myography. Surg Gynecol Obstet 1953;97:677–92.
21. Singh K, Jakab M, Reid WM, Berger LA, Hoyte L. Three-dimensional magnetic resonance imaging assessment of levator ani morphologic features in different grades of prolapse. Am J Obstet Gynecol 2003;188:910–5.
22. Smith AR, Hosker GL, Warrell DW. The role of partial denervation of the pelvic floor in the aetiology of genitourinary prolapse and stress incontinence of urine: a neurophysiological study. Br J Obstet Gynecol 1989;96:24–8.
23. Gilpin SA, Gosling JA, Smith AR, Warrell DW. The pathogenesis of genitourinary prolapse and stress incontinence of urine. A histological and histochemical study. Br J Obstet Gynaecol 1989;96:15–23.
24. Weidner AC, Barber MD, Visco AG, Bump RC, Sanders DB. Pelvic muscle electromyography of levator ani and external anal sphincter in nulliparous women and women with pelvic floor dysfunction. Am J Obstet Gynecol 2000;183:1390–9.
25. Hoyte L, Schierlitz L, Zou K, Flesh G, Fielding J. Two and 3-dimensional MRI comparison of levator ani structure, volume, and integrity in women with stress urinary incontinence and prolapse. Am J Obstet Gynecol 2001;185:11–9.
26. Lien KC, Mooney B. DeLancey JO, Ashton-Miller JA. Levator ani muscle stretch induced by simulated vaginal birth. Obstet Gynecol 2004;103:31–40.
27. Allen RE, Hosker GL, Smith AR, Warrell DW. Pelvic floor damage and childbirth: a neurophysiological study. Br J Obstet Gynaecol 1990;97:770–9.
28. Damaser MS, Broxton-King C, Ferguson C, Kim FJ, Kerns JM. Functional and neuroanatomical effects of vaginal distention and pudendal nerve crush in the female rat. J Urol 2003;170:1027–31.
29. Lien KC, Morgan DM, Delancey JO, Ashton-Miller JA. Pudendal nerve stretch during vaginal birth: a 3D computer simulation. Am J Obstet Gynecol 2005;192:1669–76.
30. Vardy MD, Lindsay R, Scotti RJ, Mikhail M, Richart RM, Nieves J, et al Short-term urogenital effects of raloxifene, tamoxifen, and estrogen. Am J Obstet Gynecol 2003;189:81–8.
31. 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.