Effects of Oxytocin for Induction and Augmentation of Labor on Pelvic Floor Symptoms and Support in the Postpartum Period

Abstract

Objective:

To determine whether oxytocin for induction or augmentation of labor impacts the incidence or persistence of pelvic floor symptoms and support 5–10 weeks after first vaginal delivery.

Methods:

Participants in this prospective cohort study were nulliparous women ≥ 18 years that delivered vaginally at ≥ 37 weeks gestation, completed the Epidemiology of Prolapse and Incontinence Questionnaire (EPIQ) and the Pelvic Organ Prolapse Quantification examination (POP-Q) in third trimester and 5–10 weeks postpartum. We compared the incidence and persistence of symptomatic EPIQ domains and worse vaginal support (maximal vaginal descent ≥ 0 cm) between women who received oxytocin to those that did not (with or without prostaglandin or mechanical methods in both groups). We performed modified binomial regression to calculate adjusted relative risks of each outcome with 95% confidence intervals.

Results:

Mean age of the 722 participants was 28.3 (SD 5.2) years; 20% were Hispanic. There were no significant differences according to oxytocin exposure in either incidence or persistence of symptomatic EPIQ domains or worse vaginal support. We found similar results in sensitivity analyses comparing women who received oxytocin as the sole pharmacologic agent to women who received no pharmacologic agent. After adjusting for demographic and obstetric factors associated with incidence and persistence of symptoms and support, oxytocin exposure continued to have no effect.

Conclusions:

Oxytocin during labor does not significantly increase the risks for the incidence or persistence of pelvic floor symptoms or worse vaginal support in the early postpartum period, though power for less frequent outcomes was limited.

Single sentence summary:

Oxytocin during labor does not significantly increase the risks for the incidence or persistence of pelvic floor symptoms or worse vaginal support in the early postpartum period.

INTRODUCTION

Pelvic floor disorders (PFDs) are common, burdensome, expensive, and the number of affected women is predicted to rise.^1–8 Therefore, identifying risk factors is important. This is made difficult because of the long-time lag between risk exposure and disease, often decades. However, pelvic floor symptoms and signs in the early postpartum period may be markers for long-term PFDs. For example, published data on stress urinary incontinence (SUI) consistently show that SUI during pregnancy increases the risk for postpartum SUI and postpartum SUI increases the risk for long-term SUI.^1,2,5,9–13

Well-established factors that increase the risk for PFDs include vaginal childbirth, obesity, age, and parity.^{1–3,5,7,10–12,14–25} When comparing women delivered vaginally to those delivered by cesarean, there is strong evidence that cesarean delivery is protective, particularly in terms of pelvic organ prolapse (POP).^{1,2,5,12,16,17,19,20,22,24,26,27} Data are mixed about the effects of some childbirth factors on PFD risk, including episiotomy, spontaneous perineal laceration, prolonged second stage of labor, and operative assisted vaginal delivery.^{1–3,6,12,15,16,20–22,27–34}

Amongst women undergoing first vaginal childbirth, most studies are in agreement that forceps deliveries increase risk of POP and urinary incontinence (UI), as well as anal sphincter rupture.^{1–3,5,6,15,20,21,28–30,32,35–37} However, the U.S. trend towards forceps use has decreased substantially, from 5.11% of all births in 1990 to 0.56% in 2015, making this risk factor for PFDs less relevant.³⁸ Data about the effect of augmentation and induction of labor on PFDs are more sparse and with mixed results, and most existing research is limited by not accounting for forceps delivery, birth weight, or length of second stage.^14,29,39,40

Labor induction for singleton births increased in the U.S. for nearly 20 years, reaching a high of 23.8% in 2010 with slight declines since then.⁴¹ However, labor induction may again increase given the recently released results of the ARRIVE randomized clinical trial in which low-risk nulliparous women randomized to elective induction of labor at 39 weeks’ gestation had improved neonatal and maternal outcomes, including decreased cesarean delivery rates, compared to women managed expectantly.⁴² Thus, the question of whether labor augmentation and induction are associated with pelvic floor symptoms, in an era of diminishing forceps use and at a gestational age associated with lower birth weight than induction in post-term pregnancy, assumes a fresh clinical relevance.

Synthetic oxytocin is the most commonly used medication on Labor and Delivery units and effectively induces labor.^43–45 Anecdotally, some women believe that synthetic oxytocin causes more painful and difficult labor, as well as PFD symptoms. In a recent non-randomized trial, pelvic floor muscle electrical activity was greater during uterine contractions in term pregnant women who received oxytocin compared to those in spontaneous labor.⁴⁶ The authors postulated that contracted pelvic floor muscles may produce counter-pressure against expulsive uterine contractions, potentially leading to obstetrical injuries.

Therefore, our aims were to determine whether oxytocin for labor augmentation or induction increases the risk for incident (primary aim) or persistent (secondary aim) pelvic floor symptoms and support between third trimester and 5–10 weeks postpartum in a population of primiparous women delivered vaginally.

MATERIAL AND METHODS

This is an ancillary prospective cohort study to an ongoing study designed to assess the impact of postpartum physical activity, intra-abdominal pressure, and fitness on pelvic floor symptoms and support one year postpartum.⁴⁷ The study was approved by the Institutional Review Boards at participating sites. All participants completed written informed consent. This ancillary study includes women that completed their 5–10 week postpartum visit by 09-24-18.

Participants, nulliparous women 18 years or older in the third trimester of pregnancy who spoke either English or Spanish, were recruited during antenatal visits at one of seven sites located in the Salt Lake Valley, including 3 university clinics, 2 private hospitals and 2 community health clinics directed at low-income women. Methods for the parent study have been published.⁴⁷ Exclusion criteria for the parent study included being unable to ambulate without assistance, prior surgery for pelvic floor disorders, connective tissue disorders related to POP, living more than 60 miles from the primary study site, and planning to move before one year postpartum. Because the aim of the parent study is focused on women delivered vaginally at term, participants subsequently delivered by cesarean and/or before 37 weeks gestation were excluded from further participation.⁴⁷

Study visits relevant to this ancillary study were conducted during the third trimester and at 5–10 weeks postpartum. At each visit, participants completed questionnaires administered via REDCap and a physical examination.^47,48 Questionnaires included demographic and other characteristics, Epidemiology of Prolapse and Incontinence Questionnaire (EPIQ), Incontinence Severity Index, and Rapid Assessment of Physical Activity (RAPA).^48–50 Trained and certified clinical coordinators conducted the physical examination which included weight, height, and the Pelvic Organ Prolapse Quantification system (POP-Q), a reproducible assessment tool for vaginal descent.^51,52 Coordinators conducted the POP-Q examination with participants in the lithotomy position with head and back elevated to 45 degrees and during strain, with the exception of total vaginal length. Clinical coordinators abstracted delivery information from the electronic medical record.

Outcomes included dichotomous measures of pelvic floor support and symptomatic EPIQ domains. We characterized support using maximal vaginal descent (MVD), the most positive value of Ba (anterior vaginal wall), Bp (posterior vaginal wall) or C (cervix), and dichotomized the support outcome as MVD ≥ 0 cm (at or below the hymen) versus < 0 cm (above the hymen) as this represents the level at which more women become clinically symptomatic.^53,54

The EPIQ assesses symptoms in six domains: SUI, overactive bladder (OAB), anal incontinence, vaginal bulge, defecation dysfunction, and voiding dysfunction/pelvic pain. We considered a domain to be positive if the response to at least one question within the domain was positive with the exception of the OAB domain. We previously reported that during pregnancy 95% of women were categorized as having OAB during the third trimester, largely because most women endorsed nocturia.⁵⁵ As the aim of this current study is focused on the incidence of symptom domains, we elected to study only the symptom of urinary urgency incontinence (UUI) from the OAB domain, as this was uncommon during pregnancy.

We grouped results from the RAPA into 3 categories: Sedentary (never/rarely do physical activity), underactive (light or moderate physical activity < 5 days/week), and active (moderate activity ≥5 days/week or vigorous activity ≥3 days/week). Given the potential link between connective tissue disorders and POP, we also queried women about the presence of varicose veins, easy bruisability and stretch marks during pregnancy.^56,57

Methods of Analysis

We defined incidence as a positive EPIQ domain (or positive UUI in the case of the OAB domain) or MVD ≥ 0 cm at 5–10 weeks, in women negative for the specific domain or MVD < 0 cm in the third trimester; and persistence, as a positive domain or MVD ≥ 0 cm at both time-points. We compared the incidence and persistence of each EPIQ domain according to oxytocin exposure using relative risk and 95% confidence intervals. Primary analyses compared 313 women who received oxytocin to 379 women who received no oxytocin, both with or without prostaglandin or mechanical methods. Planned sensitivity analyses compared 207 women who received oxytocin as the sole pharmacologic agent to 327 women who received no pharmacologic agent, both with or without mechanical methods.

Because of the large number of outcomes, we characterized demographic and obstetric risk factors other than oxytocin for participants in three groups only: 1) at least one incident symptom in any domain (vs 0), 2) at least one persistent symptom in any domain (vs 0), and 3) incident MVD ≥ 0 cm (vs < 0 cm). We used chi-squared tests, Fisher’s exact test, or independent samples t-tests, as appropriate. We modeled the effects of oxytocin on incidence and persistence of EPIQ domains with sufficient sample using modified binomial regression, with generalized estimating equations (GEE) to correct the variance of the risk ratio.⁵⁸ Since a directed acyclic graph (DAG) revealed no colliders, we adjusted individual models of incidence and persistence for selected risk factors identified in the literature, and at least marginally significant on the univariate analysis described above, when cell sizes permitted.⁵⁹ Thus, we adjusted models for the effects of oxytocin on incident EPIQ domains for forceps delivery, quartiles of birthweight, and quartiles of time since delivery, but not anal sphincter laceration as this falls on the causal pathway between birthweight or forceps and the outcomes. We adjusted models for the effects of oxytocin on persistent EPIQ domains for family history of prolapse or incontinence, ethnicity, quartiles of age, and quartiles of BMI. We adjusted the model for incident MVD ≥ 0 cm for breastfeeding, quartiles of education, and quartiles of time since delivery. We used SAS v 9.4 for analyses.

Adjustment for multiple testing

The tables display unadjusted p-values. However, we did not declare statistical significance within a symptom burden domain unless the p-value was less than 0.05/6=0.0083 for incident symptoms, reflecting 6 symptom domains; and p<0.05/4=0.0125, for persistent symptoms, reflecting 4 domains with cell sizes adequate for analysis. Cell sizes were inadequate for persistent MVD >0 cm, persistent prolapse symptoms, or persistent UUI. When a variable in quartiles was statistically nonsignificant, we further divided alpha by 3, reflecting tests of individual quartiles vs the reference quartile.

Adequacy of sample size

Sample size was constrained by the numbers available from the larger study. This information can be found in a methods paper for that study.⁴⁷ For this ancillary analysis, no sample size calculation formulae are available for modified binomial regression, so they were approximated using the z-test of two independent proportions. We assumed 38.4% frequency of oxytocin with or without prostaglandins or mechanical methods and outcome rates in the non-exposed the same as in Table 2. The sample size was sufficient with at least 80% power to detect a risk ratio of 2.0 or higher for all incident outcomes except bulge when alpha=0.0083; and for all persistent outcomes except bulge and MVD >0, when alpha=0.0125.

Table 2:

Pelvic floor symptoms and support between third trimester and 5–10 weeks postpartum

Pelvic Floor Symptom Domains and Support	Prevalence 3rd trimester n (%)	Prevalence 5–10 weeks postpartum n (%)	Incidence n (%)	Persistence n (%)	Never n (%)
Pelvic organ prolapse	43/720 (5.97%)	44/691 (6.37%)	38/677 (5.61%)	6/43 (13.95%)	609/689 (88.39%)
Stress urinary incontinence	456/721 (63.25%)	313/695 (45.04%)	63/265 (23.77%)	250/456 (54.82%)	193/690 (27.97%)
Urgency urinary incontinence	74/718 (10.31%)	133/696 (19.11)	101/644 (15.68%)	31/74 (41.89%)	519/691 (75.11%)
Pain/voiding dysfunction	252/721 (34.95%)	146/695 (21.01%)	75/469 (15.99%)	71/252 (28.17%)	374/695 (53.81%)
Anal incontinence	188/720 (26.11%)	157/695 (22.59%)	74/532 (13.91%)	83/188 (44.15%)	437/693 (63.06%)
Defecating dysfunction	355/721 (49.24%)	293/695 (42.16%)	92/366 (25.14%)	200/355 (56.34%)	257/694 (37.03%)
Maximal vaginal descent ≥ 0 cm	24/722 (3.32%)	106/716 (14.80%)	98/698 (14.04%)	8/24 (33.33%)	596/716 (83.24%)

Incidence = Symptom present postpartum/ Symptom absent in third trimester

Persistence =Symptom present postpartum/Symptom present in third trimester

RESULTS

Of 1078 participants enrolled, 825 delivered vaginally at term. After excluding those that withdrew after delivery or had incomplete delivery data, 722 (87.5%) were eligible for this analysis (Figure 1). Participant characteristics are summarized in Table 1. Mean age was 28.3 years (SD 5.2). One-fifth of participants were Hispanic. The mean interval between delivery and the postpartum visit was 7.4 (SD 1.4) weeks.

Figure 1:

Participant flow.

Table 1:

Participant demographics

	n	%
Demographics
Age, years (n=722)
18 to <25	198	27.4
25 to ≤35	448	62.1
>35	76	10.5
Race/Ethnicity (n=722)
Hispanic	144	19.9
Non-Hispanic	578	80.1
Education (n=721)
High school or less	108	15.0
Some college/completed college	410	56.9
Graduate or professional degree	203	28.2
Body mass index kg/m2 (n=715)
<18.5	4	0.6
18.5 to <25	332	46.4
25 to <30	233	32.6
30 to <35	91	12.7
≥35	55	7.7
Family history of POP/UI* (N=677)
No	424	62.6
Yes	253	37.4
Delivery Variables
Labor induction (n=721)
No	468	64.9
Yes	253	35.1
Labor augmentation (n=721)
No	525	72.8
Yes	196	27.2
Categories of labor induction/augmentation (n=722)
Oxytocin only	111	15.4
Oxytocin plus mechanical	96	13.3
Oxytocin plus prostaglandin	36	5.0
Oxytocin plus prostaglandin plus mechanical	70	9.7
Prostaglandin only	21	2.9
Prostaglandin plus mechanical	31	4.3
Mechanical only	62	8.6
Spontaneous	265	36.7
Others	30	4.2
Infant birthweight baby (grams) (N=707)
<2500 grams	14	1.9
2500–3000 grams	139	19.7
3000–3500 grams	339	48.0
3500–4000 grams	186	26.3
>4000 grams	29	4.1
Duration of second stage labor (minutes) (n=544)
0–60 minutes	210	38.6
61–120 minutes	177	32.5
120–180 minutes	81	14.9
>180 minutes	76	14.0
Epidural Anesthesia (n=722)
No	125	17.3
Yes	597	82.7
Episiotomy (n=714)
No	652	91.3
Yes	62	8.7
Forceps (n=722)
No	676	93.6
Yes	46	6.4
Vacuum (n=722)
No	708	98.1
Yes	14	1.9
Anal sphincter laceration (n=710)
No	683	96.2
Yes	27	3.8
	n	%
5–10 Weeks Postpartum
Work status (n=695)
Work in home	578	83.1
Work out of home	117	16.9
Heavy lifting or heavy work (n=694)
No	683	98.4
Yes	11	1.6
Physical activity (n=722)
Sedentary
No	703	97.4
Yes	19	2.6
Underactive
No	229	31.7
Yes	493	68.3
Active
No	549	76
Yes	173	24
Chronic cough (n=695)
No	685	98.6
Yes	10	1.4
Urinary tract infections since delivery (n=694)
No	660	95.1
Yes	34	4.9
Performs pelvic floor muscle exercises (n=694)
No	352	50.7
Yes	342	49.3
Breastfeeding (n=694)
No	85	12.3
Yes	609	87.8
Oral contraception (n=689)
No	578	83.9
Yes	111	16.1
Other hormonal methods of contraception (n=683)
No	458	67.1
Yes	225	32.9
Antidepressant use (n=683)
No	600	87.9
Yes	83	12.2
Signs of connective tissue (CT) fragility in third trimester
Varicose Veins (n=721)
No	648	89.9
Yes	73	10.1
Bruise easily (n=720)
No	562	78.1
Yes	158	21.9
Stretch marks (n=667)
No	331	46.2
Yes	386	53.8
At least one sign of CT fragility (n=717)
No	235	32.8
Yes	482	67.2

^*POP/UI = Pelvic Organ Prolapse/Urinary Incontinence

Labor was induced in 253 (35.1%) women and augmented in 196 (27.2%). Mean birthweight was 3309 g (SD 396) grams and median second stage 78.5 (IQR 88) minutes.

Trajectories of EPIQ domains (or UUI) and MVD≥0 are summarized in Table 2. When comparing oxytocin exposure versus no oxytocin exposure, there were no significant differences in either the incidence or persistence of any of the EPIQ domains (or UUI) or in the incidence or persistence of MVD ≥ 0 cm (Table 3). These findings were consistent in the sensitivity analyses excluding women that were exposed to prostaglandins.

Table 3:

Univariate analysis of risk of the incidence and persistence of pelvic floor symptoms and support by oxytocin exposure

Pelvic Floor Symptom Domains and Support	Incidence among women exposed to oxytocin	Incidence among women not exposed to oxytocin	Oxytocin exposure and risk of incident symptoms/sign: RR (95% CI)	Persistence among women exposed to oxytocin	Persistence among women not exposed to oxytocin	Oxytocin exposure and risk of persistent symptoms/sign: RR (95% CI)
Pelvic organ prolapse	18/281 (6.4%)	19/341 (5.6%)	1.15 (0.62, 2.15)	1/23 (4.4%)	4/18 (22.2%)	0.20 (0.02, 1.60)
Stress urinary incontinence	27/113 (23.9%)	34/133 (25.6%)	0.93 (0.60, 1.45)	114/192 (59.4%)	124/230 (53.9%)	1.10 (0.93, 1.30)
Urgency urinary incontinence	49/269 (18.2%)	46/328 (14.0%)	1.30 (0.90, 1.88)	20/36 (55.6%)	11/33 (33.3%)	1.67 (0.95, 2.93)
Pain/voiding dysfunction	30/195 (15.3%)	44/237 (18.6%)	0.83 (0.54, 1.27)	29/110 (26.4%)	38/126 (30.2%)	0.87 (0.58, 1.32)
Anal incontinence	33/219 (15.1%)	37/274 (13.5%)	1.12 (0.72, 1.72)	41/86 (47.7%)	38/88 (43.2%)	1.10 (0.80, 1.53)
Defecating dysfunction	38/150 (25.3%)	50/182 (27.5%)	0.92 (0.64, 1.32)	87/155 (56.1%)	106/181 (58.6%)	0.96 (0.80, 1.15)
Maximal vaginal descent ≥ 0 cm	35/300 (11.7%)	60/364 (16.5%)	0.71 (0.48, 1.04)	5/10 (50%)	3/12 (25.0%)	2.00 (0.63, 6.38)

Incidence = Symptom present postpartum/ Symptom absent in third trimester

Persistence =Symptom present postpartum/Symptom present in third trimester

Univariate analyses of potential risk factors for the incidence or persistence of symptoms in at least one EPIQ domain are summarized in Table 4. Additionally (data not shown), univariate analysis revealed significant associations between incident MVD ≥ 0 cm and education (RR 1.90 (95% CI 1.12, 3.22)) for ≤ high school vs graduate/professional degree), breastfeeding (RR 0.59 (95% CI 0.38, 0.94)) for women breastfeeding vs not, and time since delivery (RR 1.23 (95% CI 1.09, 1.39)) for each additional day. The mean (SD) number of weeks between delivery and the postpartum evaluation was 7.83 (SD 1.51) for women with incident MVD ≥ 0 cm vs 7.30 (SD 1.38) for those without (p=0.0006).

Table 4:

Univariate analysis of risk of the incidence and persistence of at least one pelvic floor symptom by demographic, delivery, and other variables

	≥ 1 Incident pelvic floor symptom (n)	(%)	p-value	RR (95% CI)	≥ 1 Persistent pelvic floor symptom (n)	(%)	p-value	RR (95% CI)
Demographics
Age, years
18 to <25 (ref.)	85	45.70			97	52.15
25 to ≤35	202	46.54	0.85	1.02 (0.85, 1.23)	268	61.75	0.03	1.18 (1.01, 1.38)
>35	32	43.24	0.72	0.95 (0.70, 1.28)	50	67.57	0.03	1.30 (1.05, 1.60)
Race/Ethnicity
Non-Hispanic (ref.)	254	45.52			361	64.70
Hispanic	65	47.79	0.63	1.05 (0.87, 1.28)	54	39.71	<0.0001	0.61 (0.49, 0.77)
Education
High school or less	42	42.42	0.34	0.89 (0.67, 1.15)	47	47.47	0.004	0.73 (0.58, 0.92)
Some college/completed college	180	45.57	0.54	0.94 (0.79, 1.13)	239	60.51	0.31	0.93 (0.82, 1.06)
Graduate or professional degree (ref.)	96	48.24			129	64.82
Body mass index kg/m 2
<18.5	3	75.00	0.63	1.51 (0.85, 2.69)	0	0	0.02	not applicable
18.5 to <25 (ref.)	160	49.69			184	57.14
25 to <30	95	42.99	0.12	0.87 (0.71, 1.04)	136	61.54	0.31	1.08 (0.94, 1.24)
30 to <35	39	43.44	0.29	0.87 (0.67, 1.13)	58	64.44	0.21	1.13 (0.94, 1.35)
≥35	19	38.00	0.12	0.76 (0.53, 1.11)	34	68.00	0.15	1.19 (0.96, 1.47)
Family history of POP/UI *
No (ref.)	186	45.93			209	51.60
Yes	115	46.94	0.80	1.02 (0.86, 1.21)	182	74.29	<0.0001	1.44 (1.28, 1.62)
Delivery Variables
Labor induction
No (ref.)	209	46.14			257	56.73
Yes	110	45.64	0.90	0.99 (0.83, 1.17)	158	65.56	0.02	1.16 (1.02, 1.31)
Labor augmentation
No (ref.)	231	45.74			305	60.40
Yes	88	46.56	0.85	1.02 (0.85, 1.22)	110	58.20	0.60	0.96 (0.84, 1.11)
Infant birthweight baby (grams)
<2500 grams	8	57.14	0.08	1.70 (1.02, 2.84)	9	64.29	0.42	1.21 (0.80, 1.85)
2500–3000 grams (ref.)	45	33.58			71	52.99
3000–3500 grams	146	44.65	0.03	1.33 (1.02, 1.74)	199	60.86	0.12	1.15 (0.96, 1.38)
3500–4000 grams	96	54.24	0.0003	1.62 (1.23, 2.12)	107	60.45	0.19	1.14 (0.94, 1.39)
>4000 grams	17	62.96	0.0042	1.88 (1.29, 2.73)	17	62.96	0.34	1.19 (0.85, 1.65)
Duration of second stage labor (minutes)
0–60 minutes	84	41.18	0.23	0.87 (0.69, 1.09)	110	53.92	0.18	0.87 (0.74, 1.06)
61–120 minutes (ref.)	81	47.37			104	60.82
120–180 minutes	46	58.23	0.11	1.23 (0.96, 1.57)	50	63.29	0.71	1.04 (0.85, 1.28)
>180 minutes	31	43.06	0.538	0.91 (0.67, 1.24)	42	58.33	0.72	0.96 (0.76, 1.21)
Epidural Anesthesia
No (ref.)	52	43.70			70	58.82
Yes	267	46.43	0.59	1.06 (0.85, 1.33)	345	60.00	0.81	1.02 (0.87, 1.20)
Episiotomy
No (ref.)	286	45.47			370	58.82
Yes	30	51.72	0.36	1.14 (0.87, 1.48)	41	70.69	0.08	1.20 (1.01, 1.44)
Forceps
No (ref.)	294	45.09			386	59.20
Yes	25	59.52	0.07	1.32 (1.01, 1.72)	29	69.05	0.21	1.17 (0.94, 1.44)
Vacuum
No (ref.)	312	45.81			409	60.06
Yes	7	53.85	0.57	1.18 (0.70, 1.96)	6	46.15	0.31	0.77 (0.43, 1.39)
Anal sphincter laceration
No (ref.)	294	44.75			394	59.97
Yes	20	76.92	0.0012	1.72 (1.37, 2.16)	15	57.69	0.82	0.96 (0.69, 1.34)
5–10 Weeks Postpartum
Work status
Less than 30 hours per week	18	36.73	0.27	0.78 (0.50, 1.22)	29	59.18	0.35	0.87 (0.66, 1.16)
More than 30 hours per week (ref.)	32	47.06			46	67.65
Other	269	46.62	0.95	0.99 (0.76, 1.29)	340	58.93	0.17	0.87 (0.73, 1.04)
Heavy lifting or heavy work
No (ref.)	311	45.6			408	59.82
Yes	7	63.64	0.23	1.40 (0.87, 2.20)	7	63.64	0.80	1.07 (0.68, 1.67)
Physical activity
Sedentary
No (ref.)	310	45.95			405	60.00
Yes	9	47.37	0.90	1.03 (0.64, 1.67)	10	52.63	0.52	0.88 (0.57, 1.35)
Underactive
No (ref.)	100	49.75			108	53.73	0.04	1.16 (1.00, 1.34)
Yes	219	44.42	0.20	0.89 (0.75, 1.06)	307	62.27
Active							0.22	0.91 (0.79, 1.06)
No (ref.)	232	44.44			319	61.11
Yes	87	50.58	0.16	1.14 (0.95, 1.36)	96	55.81
Chronic cough
No (ref.)	315	46.05			408	59.65
Yes	4	40.00	0.76	0.87 (0.40, 1.86)	7	70.00	0.51	1.17 (0.78, 1.77)
Urinary tract infections since delivery
No (ref.)	300	45.52			392	59.48
Yes	18	52.94	0.40	1.16 (0.84, 1.61)	23	67.65	0.34	1.14 (0.89, 1.45)
Performs pelvic floor muscle exercises
No (ref.)	158	45.01			215	61.25
Yes	161	47.08	0.59	1.05 (0.89, 1.23)	200	58.48	0.46	0.95 (0.85, 1.08)
Breastfeeding
No (ref.)	34	40.00			44	51.76
Yes	284	46.71	.25	1.17 (0.89, 1.54)	370	60.86	0.11	1.18 (0.95, 1.46)
Oral contraception **
No (ref.)	263	45.50			338	58.48
Yes	53	48.18	0.61	1.06 (0.86, 1.31)	74	67.27	0.09	1.15 (0.99, 1.33)
Other hormonal methods of contraception ***
No (ref.)	207	45.30			276	60.39
Yes	108	48.00	0.51	1.06 (0.89, 1.26)	132	58.67	0.67	0.97 (0.85, 1.11)
Antidepressant use
No (ref.)	280	46.74			355	59.27
Yes	34	40.96	0.32	0.88 (0.67, 1.15)	54	65.06	0.31	1.10 (0.93, 1.30)
Association between categorical time since delivery to 6 week visit and maximal vaginal descent incidence
4 weeks to 6 weeks (ref.)	10	7.58
6 weeks to 9 weeks	64	13.59	0.06	1.79 (0.95, 3.39)
9 weeks to 12 weeks	24	21.24	0.002	2.80 (1.40, 5.61)
Signs of connective tissue (CT) fragility in third trimester
Varicose Veins
No (ref.)	287	46.07			373	59.87
Yes	32	45.71	0.96	0.99 (0.76, 1.30)	42	60.00	0.98	1.00 (0.82, 1.23)
Bruise easily
No (ref.)	242	44.90			321	59.55
Yes	76	49.67	0.30	1.11 (0.92, 1.33)	93	60.78	0.78	1.02 (0.88, 1.18)
Stretch marks
No (ref.)	160	50.00			181	56.56
Yes	158	42.82	0.06	0.86 (0.73, 1.01)	230	62.33	0.12	1.10 (0.97, 1.25)
At least one sign of CT fragility
No (ref.)	111	48.47			131	57.21
Yes	206	44.78	0.36	0.92 (0.78, 1.09)	281	61.09	0.33	1.07 (0.93, 1.22)

^*POP/UI = Pelvic Organ Prolapse/Urinary Incontinence

^**Oral contraception: all pill forms of birth control including estrogen and progesterone containing and progesterone only.

^***Other hormonal methods of birth control: including progesterone-releasing intrauterine devices, progesterone implants, progesterone injectables, hormonal patch, and hormonal vaginal ring.

In adjusted models for symptom and support incidence or persistence, oxytocin was never statistically significant (p>0.05 for all models). Other factors that were statistically significant predictors for outcomes are summarized as follows. Quartiles of time since delivery increased the risk for incident defecation dysfunction (p=0.0082) with a strong log-linear trend for higher risk of symptoms later in time. Higher birthweight increased the risk for incident pain/voiding dysfunction (RR 3.00 (95% CI1.50, 5.82); p=0.0083)) for the highest vs the second quartile (reference). Use of forceps increased the risk for incident anal incontinence symptoms (RR 3.45 (95% CI 1.97, 6.04); p<0.0001)). Self-reported family history of POP or UI increased the risks of persistent SUI (RR 1.79 (95% CI 1.46, 2.20)) and persistent pain/voiding dysfunction (RR 2.53 (95% CI (1.54, 4.16)); p<0.001 in each model. Compared to non-Hispanic ethnicity, Hispanic ethnicity decreased the risk for persistent defecation dysfunction (RR 0.57 (95% CI 0.36, 0.88), p=0.0036).

DISCUSSION

In this prospective study, we found no significant differences in either incidence or persistence of symptomatic EPIQ domains or MVD≥0 cm, in primiparous women exposed to oxytocin compared to those that were not. Oxytocin exposure continued to have no effect in sensitivity analyses restricted to women that received no other pharmacologic agents for labor induction or augmentation or in models adjusted for demographic and obstetric factors associated with incidence and persistence of symptoms and support.

Our results are in agreement with two studies that found minimal effect of labor induction on the incidence of postpartum SUI and another that concluded that oxytocin did not increase the risk of levator ani muscle avulsion.^6,40,60 In contrast, in a retrospective cohort study women exposed to a single dose of oxytocin had 1.85 odds (95% 1.04–3.31) and those exposed to two doses 3.07 odds (95% CI 1.01–9.31) of UI compared to women with no exposure.²⁷ Another study concluded that oxytocin augmentation nearly doubled the odds of anal incontinence.⁶¹ Two additional studies found that the risk for pelvic floor disorders increased with each labor induction.^14,62

Strengths of this study include its prospective nature, large sample size, high follow-up rate and use of validated instruments. Coordinators conducting POP-Q examinations were trained and certified before beginning.⁴⁷

This study has limitations. Given the non-randomized nature of the design, indication bias may have influenced our results. Our results may not be generalizable to racial groups other than Caucasian or to multiparous women. We had limited power to investigate less frequent outcomes. We did not assess dose or duration of oxytocin use, and others have found that maternal and fetal complications linked to oxytocin may be dose dependent and influenced by duration of exposure.^63,64 We focused on oxytocin as it is commonly drug used, and one mechanistic study suggested potential biological plausibility for its effect on the pelvic floor.^43,45,46 Other methods of labor induction and augmentation may have different effects. The follow-up is short, but this time interval was chosen because women are more likely to demonstrate symptoms at 5–10 weeks postpartum than later in the postpartum year and thus, if an effect is present it is likely to be present early after delivery. If there is a delayed effect of oxytocin on pelvic floor health, this short-term analysis would not demonstrate it.

Given that the ARRIVE trial concluded that elective induction at 39 weeks’ gestation is safe, elective induction of labor may become more popular in the future. When discussing benefits and risks of elective induction with patients, clinicians can use our results to add information about pelvic floor risks. Based on our results, the use of oxytocin for labor induction and augmentation does not appear to significantly increase the risk of new onset or persistence of pelvic floor symptoms or worsened vaginal support postpartum, amongst nulliparous women ultimately delivered vaginally.