Background
Pelvic floor disorders (including urinary and anal incontinence and pelvic organ prolapse) are associated with childbirth. Injury to the pelvic floor muscles during vaginal childbirth, such as avulsion of the levator ani muscle, is associated with weaker pelvic floor muscle strength. As weak pelvic floor muscle strength may be a modifiable risk factor for the later development of pelvic floor disorders, it is important to understand how pelvic floor muscle strength affects the course of pelvic floor disorders over time.
Objective
To investigate the association between pelvic floor muscle strength and the incidence of pelvic floor disorders, and to identify maternal and obstetrical characteristics that modify the association.
Materials and Methods
This is a longitudinal study investigating pelvic floor disorders after childbirth. Participants were recruited 5–10 years after their first delivery and were assessed for pelvic floor disorders annually for up to 9 years. Stress incontinence, overactive bladder, and anal incontinence were assessed at each annual visit using the Epidemiology of Prolapse and Incontinence Questionnaire. Pelvic organ prolapse was assessed on physical examination, and was defined as descent of the vaginal walls or cervix beyond the hymen during forceful Valsalva. The primary exposure of interest was pelvic floor muscle strength, defined as the peak pressure during a voluntary pelvic muscle contraction (measured with a perineometer). The relationship between pelvic floor muscle strength and the cumulative incidence (time to event) of each pelvic floor disorder was evaluated using lognormal models, stratified by vaginal vs cesarean delivery. The relative hazard for each pelvic floor disorder (among those women free of the disorder at enrollment and thus more than 5–10 years from first delivery), was estimated using semiparametric proportional hazard models as a function of delivery mode, pelvic floor muscle strength, and other covariates.
Results
Of 1143 participants, the median age was 40 (interquartile range, 36.6–43.7) years, and 73% were multiparous. On perineometry, women with at least 1 vaginal delivery were more likely to have a low peak pressure, defined as <20 cm H 2 O (243 of 588 women with at least 1 vaginal delivery vs 107 of 555 women who delivered all of their children by cesarean delivery, P < .001). Among women who had at least 1 vaginal delivery, a pelvic floor muscle strength of <20 cm H 2 O was associated with a shorter time to event for stress incontinence (time ratio, 0.67; 95% confidence interval, 0.50–0.90), overactive bladder (time ratio, 0.67; 95% confidence interval, 0.51–0.86), and pelvic organ prolapse (time ratio, 0.76; 95% confidence interval, 0.65–0.88). No such association was found among women who delivered all of their children by cesarean delivery. Among women with at least 1 vaginal delivery and considering only pelvic floor disorders that developed during study observation (5–10 years after the first delivery), and controlling for maternal characteristics (body mass index and genital hiatus), women who had a peak pressure of <20 cm H 2 O had hazard ratios (relative to ≥20 cm H 2 O) of 1.16 (95% confidence interval, 0.74–1.81) for stress incontinence, 1.27 (95% confidence interval, 0.78–2.05) for overactive bladder, and 1.43 (95% confidence interval, 0.99–2.07) for pelvic organ prolapse. Among women who delivered all of their children by cesarean delivery, there was no association between muscle strength and relative hazard of pelvic floor disorders when controlling for maternal characteristics.
Conclusion
After vaginal delivery, but not cesarean delivery, the cumulative incidence of pelvic organ prolapse, stress incontinence, and overactive bladder is associated with pelvic muscle strength, but the associations attenuate when adjusting for genital hiatus and body mass index.
Pelvic floor disorders (PFDs), including urinary incontinence, anal incontinence, and pelvic organ prolapse, are associated with childbirth. Specifically, these disorders are strongly associated with parity and are more common after a vaginal than a cesarean delivery. Obstetric injury or impairment of levator ani muscle function may be a mechanism by which vaginal delivery contributes to the later development of prolapse and other PFDs. For example, levator ani avulsions, which are seen in 10–20% of vaginal deliveries, are associated with pelvic organ prolapse (POP). Levator ani avulsions are also associated with weaker pelvic floor muscles. This finding, along with a computer simulation model showing that pelvic muscle weakness results in progressive uterovaginal prolapse, suggests that pelvic floor muscle strength may be important in the development of PFDs.
Why was this study conducted?
To investigate the effect of pelvic floor muscle strength on the development of pelvic floor disorders up to 2 decades after childbirth.
Key findings
Among women who had at least 1 vaginal delivery, reduced pelvic floor muscle strength was associated with an increased incidence of stress urinary incontinence, overactive bladder, and pelvic organ prolapse. This association was not found among women who had all of their children by cesarean delivery.
What does this add to what is known?
This study helps to identify women who are at higher risk for the development of pelvic floor disorders after childbirth, and who may therefore benefit from preventive strategies such as strengthening of the pelvic floor muscles.
Pelvic floor muscle training is a recommended treatment option for urinary incontinence and pelvic organ prolapse, and is recommended as first-line therapy for women with stress urinary incontinence by the American College of Physicians. As pelvic floor muscle strength could thus be a potentially modifiable risk factor for PFDs, it is important to understand how pelvic floor muscle strength affects the course and progression of PFDs over time. The aims of this study were to investigate the association between pelvic floor muscle strength and the incidence of PFDs, and to identify maternal and obstetrical characteristics, if any, that modify the association. We hypothesized that weak pelvic floor muscle strength is an independent risk factor for the development of PFDs.
Materials and Methods
This research reflects 1 of the primary aims of the Mother’s Outcome After Delivery (MOAD) study, a longitudinal study of PFDs after childbirth. MOAD participants were recruited from a community hospital 5–10 years after their first delivery. Detailed recruitment methods have previously been reported. Briefly, recruitment was based on delivery mode (cesarean vs vaginal delivery), and delivery groups were matched by age and years since first delivery. Exclusion criteria (for the index birth) included the following: maternal age <15 or >50 years, delivery at <37 weeks’ gestation, placenta previa, multiple gestation, known fetal congenital anomaly, stillbirth, prior myomectomy, and abruption. Participants were enrolled between October 2008 and December 2013. Annual follow-up continued through April 2017. The study was approved by both the Johns Hopkins and Greater Baltimore Medical Center Institutional Review Boards, and all participants signed an informed consent.
Participants were seen at the research site for a baseline visit and annually thereafter for up to 9 years. At each visit, women completed questionnaires and had a physical examinaton, including a Pelvic Organ Prolapse Quantification (POP-Q) examination. Pelvic floor muscle strength was measured at the second annual visit. Those who were not present for a second annual visit were asked to participate at later visits. Examinations were performed by female pelvic medicine physicians and a research nurse, each of whom demonstrated competency prior to the study.
Exposures
The primary exposure of interest for this study was pelvic floor muscle strength, measured using the Peritron perineometer (CardioDesign, Oakleigh, Australia). The Peritron consists of a 28-mm-diameter compressible probe connected to a handheld microprocessor. Previous research has shown that measurements obtained with the Peritron are reproducible and reliable. Participants were instructed to squeeze the pelvic floor muscles as if they were trying to hold in flatus. Before placing the probe, digital palpation by the researcher confirmed proper technique after coaching as needed. With the Peritron probe in the vagina, participants were asked to contract the pelvic floor muscles as forcefully as possible and to maintain the contraction for as long as possible. The procedure was repeated after a 10-second rest. The average peak pressure (in centimeters of water) and contraction duration (in seconds) were recorded. Our protocol for perineometer measurement is based on that described by Hundley et al. Peak pressure was classified into 2 groups (<20 cm H 2 O and ≥20 cm H 2 O) based on tertiles (one-third of all participants had a peak pressure of <20 cm H 2 O). Women reporting a latex allergy were excluded, as the tubing used with this device contains latex.
Outcomes
The primary outcome was the incidence of PFDs: namely, stress incontinence (SUI), overactive bladder (OAB), anal incontinence (AI), and pelvic organ prolapse (POP). PFDs were assessed at each annual visit using validated questionnaires and a physical examination. SUI, OAB, and AI were assessed using the Epidemiology of Prolapse and Incontinence Questionnaire (EPIQ). Participants with scores above the previously validated threshold scores were considered to have that disorder. POP was assessed on physical examination using the POP-Q system and was defined as descent of the vaginal walls or cervix beyond the hymen during forceful Valsalva maneuver. At each visit, participants were asked about prior treatment for PFDs, including surgery, supervised pelvic muscle exercises, pessary use, and medications. Women who reported therapy for a specific PFD were considered to have that condition, regardless of current symptoms. Study personnel performing the POP-Q examination and the perineometry were blinded to obstetrical history and current symptoms.
Additional covariates
Maternal and obstetric characteristics were obtained from review of obstetric records as well as information obtained at each annual visit. Women were classified into 2 obstetric categories: cesarean delivery and vaginal delivery. The cesarean delivery group included women who delivered all of their children via cesarean delivery, whereas the vaginal delivery group included women who had at least 1 vaginal delivery (spontaneous or operative vaginal delivery). The cesarean group included women who had an unlabored cesarean delivery, a cesarean delivery during active labor, or a cesarean delivery after complete cervical dilation, as prior research in this cohort has shown no difference in odds of PFDs or pelvic muscle strength in these groups. Maternal characteristics considered included race, body mass index defined as weight (kg)/height (m) , and genital hiatus. Race was self-reported at study entry. Body mass index (BMI) was measured at each annual visit. Genital hiatus (ie, the distance from the middle of the external urethral meatus to the posterior midline hymen, measured during the Valsalva maneuver) was measured annually as part of the POP-Q examination. The genital hiatus was classified into 1 of 3 groups, measured to the nearest half centimeter (≤2.5 cm, 3 cm, ≥3.5 cm).
Statistical analysis
The first aim of this study was to describe the association between pelvic floor muscle strength and the incidence (time to event) of PFDs. For each PFD, all study participants contributed data in estimating incidence so that those who developed the PFD prior to study entry contributed to the analysis as left-censored observations; those who developed the PFD during follow-up contributed as uncensored observations; and those who exited the study free of the disorder contributed to the analysis as right-censored observations. Conventional lognormal models were used to estimate the cumulative incidence of each PFD from first delivery over time, stratified by pelvic floor muscle strength (ie, <20 cm H 2 O or ≥20 cm H 2 O). As delivery mode is associated with both PFDs and pelvic muscle strength, we further stratified based on delivery mode. These models produced time ratios, measuring the magnitude by which times to events expand or contract relative to the reference group.
For the second aim, to determine which maternal and obstetrical characteristics, if any, modify the association between pelvic floor muscle strength and PFDs, semiparametric proportional hazards models were used. Women who were not disorder free at study entry were excluded from this analysis, as these cases occurred before study observation, and therefore the characteristics of the women at the time the condition arose (eg, BMI) were unknown. Five years from the first delivery was thus used as the origin, and methods allowing for staggered entries were used to account for women surviving disorder free to study entry. The independent variables included in this analysis included pelvic floor muscle strength, delivery mode, BMI, and the size of genital hiatus. Based on prior research in this cohort, which showed minimal change in pelvic floor muscle strength over time, pelvic floor muscle strength was considered to be constant over time. Delivery mode was also considered to be constant over time, as transitions between the vaginal and cesarean birth groups were infrequent in this population. The relative hazard was estimated as a function of these fixed covariates (pelvic floor muscle strength and delivery group) and the current status of the time varying measures (BMI and genital hiatus).
Data were not complete for BMI and genital hiatus at all follow-up visits (n = 229; 3.3% had missing data). Missing data were imputed with a carry forward-method for up to 1 visit (ie, if a value was missing, the value at the previous visit was used). After such imputation, BMI and genital hiatus remained unavailable in 37 (0.5%) observations.
Two-sided 95% confidence intervals (CI) were reported alongside each estimate of association (time and hazard ratios) and indicated statistical significance when intervals did not contain the null value 1. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC) and graphics were created in R version 3.4.3 (R Foundation for Statistical Computing, Vienna, Austria).
Results
A total of 1143 of the 1529 women enrolled in the MOAD Study participated in this complementary study. Those not enrolled included 297 women who did not return for a second visit, 42 women with a latex allergy, 26 who declined, and 21 who did not participate for other reasons. At perineometry, the median age was 40.0 years (interquartile range [IQR], 36.6–43.7) and the median duration from first delivery was 7.7 years (IQR, 6.8–9.6). The majority (73%) were multiparous, with 16% having had 3 or more children; 555 participants delivered all of their children by cesarean delivery, whereas 588 women had at least 1 vaginal delivery. The median peak pressure measurement at perineometry was 22.5 cm H 2 O (IQR, 13.5–34.5) among women who had at least 1 vaginal delivery and 35.5 cm H 2 O (IQR, 22.8–49.5) among women who had all of their children via cesarean delivery.
Table 1 describes characteristics of participants, comparing women with peak pressure <20 cm H 2 O vs those with peak pressure ≥20 cm H 2 O. There was no significant difference between groups based on age at first delivery, race, parity, or BMI. Participants with a peak pressure <20 cm H 2 O were more likely to have had at least 1 vaginal delivery and a genital hiatus measurement of ≥3.5cm. The number of visits was similar in the 2 groups, so the prospective data available for the analysis were comparable in the 2 groups of interest.
| Characteristic | Peak pressure a <20 cm H 2 O n = 350 (31%) | Peak pressure a ≥20 cm H 2 O n = 793 (69%) | P value |
|---|---|---|---|
| Delivery group at entry | <.001 | ||
| Cesarean only | 107 (30.6) | 448 (56.5) | |
| Vaginal | 243 (69.4) | 345 (43.5) | |
| Age at first delivery, y | .557 | ||
| <30 | 125 (35.7) | 308 (38.8) | |
| 30 to <35 | 124 (35.4) | 275 (34.7) | |
| ≥35 | 101 (28.9) | 210 (26.5) | |
| Black race | 42 (12.0) | 111 (14.0) | .408 |
| Parity at entry | .299 | ||
| 1 | 86 (24.6) | 218 (27.5) | |
| 2 | 200 (57.1) | 456 (57.5) | |
| ≥3 | 64 (18.3) | 119 (15.0) | |
| BMI, kg/m 2 , at enrollment | .066 | ||
| <25 | 183 (52.3) | 361 (45.5) | |
| 25 to <30 | 97 (27.7) | 231 (29.1) | |
| ≥30 | 70 (20.0) | 201 (25.3) | |
| Genital hiatus b at enrollment, cm | <.001 | ||
| ≤2.5 | 156 (44.6) | 503 (63.4) | |
| 3 | 69 (19.7) | 152 (19.2) | |
| ≥3.5 | 125 (35.7) | 138 (17.4) | |
| No. of follow-up visits | .503 | ||
| 2 or 3 | 52 (14.9) | 99 (12.5) | |
| 4–6 | 200 (57.1) | 456 (57.5) | |
| 7–9 | 98 (28.0) | 238 (30.0) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
