Effect of Interpregnancy Interval on the Development of Pelvic Floor Disorders

Abstract

Importance

The effect of a short interpregnancy interval (IPI) on pelvic floor disorders is unknown.

Objectives

We investigated the relationship between a short IPI and the development of stress incontinence (SUI), pelvic organ prolapse (POP), and anal incontinence (AI) in the decade after the first delivery.

Study Design

We performed a secondary analysis of the Mothers’ Outcomes After Delivery study—a prospective cohort of women recruited 5–10 years following their first delivery and followed annually between 2008 and 2018. A short IPI was defined as ≤ 18 months, calculated as the number of months between deliveries minus the length of the second pregnancy. SUI, POP, and AI were identified by annual validated questionnaires, examination, or history of treatment. Data were analyzed in SAS. Variables significant on bivariate analysis were entered into multivariable logistic regression models predicting each outcome using the generalized estimating equations approach for repeated measures.

Results

Of 1,127 women, the majority (671, 59%) never had a short IPI, 395 (35%) had 1 short IPI, and 61 (6%) had 2 or more. Within 10–15 years, 219 women (19%) reported SUI, 156 (14%) reported POP, and 251 (22%) reported AI. We found that a short IPI was not associated with SUI (P = 0.69), POP (P = 0.71), or AI (P = 0.95). When restricting the cohort to women with only nonoperative vaginal deliveries (n = 440), there remained no difference in the presence of SUI, POP, or AI (P = 0.88, 0.84, 0.78, respectively).

Conclusions

A short IPI is not associated with pelvic floor disorders in the decade following the first delivery. This should be reassuring to women at risk of PFDs who elect to become pregnant within 18 months of childbirth.

Global estimates of the prevalence of pelvic floor disorders (PFDs) range from 35% to 38%, with some reports as high as 46–50%.^1,2 In the United States alone, it has been demonstrated that approximately 25% of women older than 20 years reported living with at least 1 PFD: urinary incontinence (UI), pelvic organ prolapse (POP), and anal incontinence (AI).^3,4 Pelvic floor disorders may remain underdiagnosed and untreated, so the true prevalence of these disorders may even be higher than current estimates.⁵

Pregnancy is one of the most significant risk factors for the development of PFDs, and while the decision to have children may not be modifiable, some of the factors around pregnancy and delivery have been studied to elucidate their effect on the development of PFDs.⁶ While parity is a commonly cited risk factor, its effect on PFDs is controversial, as multiple studies have shown that multiparity does not increase the risk of POP or UI.^7–9 In contrast, the mode of delivery does play an important role in the development of PFDs. Women who have vaginal deliveries have an increased risk of developing PFDs compared with cesarean delivery, with operative vaginal deliveries carrying the greatest risk.^8,10–12 This association has been attributed to anatomic changes that occur to the pelvic floor during vaginal delivery and to the potential for additional injury to structures such as the levator ani and anal sphincter muscles from operative vaginal delivery.^13,14

The American College of Obstetricians and Gynecologists (ACOG) advises that women should avoid an interpregnancy interval (IPI) of less than 18 months due to increased risk of adverse obstetrical outcomes.¹⁵ However, there is very little information about whether PFDs are associated with the IPI, an obstetrical characteristic that may be modifiable by the patient. Studies suggest that it can take 12 months or longer for pelvic floor strength to recover after childbirth, and for some, full recovery may never occur.^16–18 Understanding whether there is an association between a short IPI and PFDs may help counsel patients about the relative risks of their family planning choices. We hypothesized that a shorter IPI may be associated with the incidence of PFDs later in life because a short IPI might prevent full postpartum pelvic floor recovery before sustaining another stressor. Therefore, the aim of this study was to investigate the relationship between a short IPI and the development of PFDs in the first 2 decades following childbirth.

MATERIALS AND METHODS

This was a secondary analysis of the Mothers’ Outcomes After Delivery (MOAD) study—a longitudinal prospective cohort study of women recruited 5–10 years following their first delivery and followed annually between 2008 and 2018.¹¹ Briefly, the primary study recruited 1,529 women from the Baltimore, Maryland area for a longitudinal cohort study 5–10 years after their first delivery (occurring between 1998 and 2008). Using hospital records, the investigators classified each birth as cesarean without labor, cesarean during active labor, cesarean after complete cervical dilation, spontaneous vaginal birth, or operative vaginal birth. The study received institutional review board approval and all patients signed informed consent.

The outcomes of interest were stress incontinence (SUI), AI, and POP. At enrollment, incontinence symptoms were assessed with the Epidemiology of Prolapse and Incontinence Questionnaire, and prolapse was assessed on examination with the Pelvic Organ Prolapse Quantification (POP-Q) system. Previously validated threshold scores were used to identify SUI and AI. Prolapse was defined as the descent of any vaginal segment to or beyond the hymen.^19,20 Participants were followed annually for up to 10 years and asked to complete the Epidemiology of Prolapse and Incontinence Questionnaire and present for a POP-Q examination. If a participant met the criteria for SUI, POP, or AI at 1 visit, they were considered to have that condition, regardless of whether the criteria were met at subsequent visits. At baseline, participants were asked about prior treatment for PFDs, including surgery. At each time point of follow-up, participants were also asked about current therapy, including medications for UI or current pessary use for treatment of prolapse. We also considered current or prior pelvic muscle exercises as treatment, but only if the program was supervised by a therapist. For the purposes of this analysis, women who reported prior surgery, prior supervised pelvic muscle exercises, or any current therapy for a specific PFD were considered to have that condition regardless of current symptoms. Patients who became pregnant and gave birth during the study continued to receive follow-up, although they were not assessed for PFDs during pregnancy. Assessments regarding the presence of PFDs resumed after 6 months postpartum.

The exposure of interest was a short IPI, defined as ≤ 18 months. The IPI was calculated as the number of months between deliveries minus the length of the second pregnancy. This analysis includes only those in the cohort who had 2 or more deliveries, irrespective of delivery type (vaginal vs cesarean), such that they would have a measurable IPI. If the gestational age at delivery of the second pregnancy was unknown, we chose to impute a below-average gestational age of 37 weeks. This was done to be conservative about the possible association between short IPIs and PFDs by underestimating the number of short IPIs. In a sensitivity analysis, participants with unknown gestational ages (and therefore, unknown IPI) were excluded.

Other characteristics were considered in multivariable analysis. Primary race was reported at study entry. Time-varying covariates, assessed at each visit, included parity, educational attainment, and tobacco use. Body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) was measured at study entry. Review of relevant obstetrical records for all deliveries allowed identification of fetal macrosomia (defined as birth weight ≥ 4,000 g), episiotomy at any delivery, anal sphincter laceration at any delivery, prolonged second stage of labor at any delivery (defined as > 2 h), and the use of vacuum or forceps for any vaginal birth. When obstetrical records were not available (eg, for deliveries that occurred at other hospitals), patient recall was used to classify obstetrical exposures.²¹

Bivariate associations between a short IPI and factors of interest were analyzed using the χ² or Fisher exact tests as appropriate for categorical factors, or the Student t-test for continuous variables. Variables significant on bivariate analysis, as well as delivery mode (which has been previously associated with SUI, POP, and AI), were included in multivariate analyses. Logistic regression predicting PFDs (SUI, AI, POP) with a short IPI were adjusted for age at first delivery, BMI, birth type category, and time-varying parity using a repeated measures generalized estimating equations model. Post hoc power was calculated for each outcome (SUI, POP, AI) using the observed sample sizes and proportions for a 2-sided test with alpha set at 0.05.

RESULTS

Of the 1,529 women in the original cohort, 402 had only 1 delivery and were excluded from this secondary analysis. As a result, 1,127 women met the criteria for analysis in this study. Most of these women had 2 (n = 838, 71%) or 3 (n = 255, 26%) deliveries. Among these women, 671 (60%) never had a short IPI, 317 (35%) had 1 short IPI, and 61 (6%) had 2 or more short IPIs. The median (IQR) IPI in those women who had a short IPI was 12.5 months (9.2, 15.1) and the median (IQR) IPI in those women who never had a short IPI was 30.1 months (18.03, 156). There were 106 participants with unknown gestational age for at least 1 pregnancy; for these pregnancies, the IPIs were imputed. Table 1 describes further characteristics of the women included in the study. The median longitudinal follow-up after first delivery was 11.2 years, with the median follow-up from first delivery to first survey at 6.4 years. On univariate analysis (Table 1), age at first delivery (P < 0.001), BMI (P = 0.005) and age at study entry (P < 0.001) were associated with a short IPI. Of the 1,127 women included in this study, 219 (19%) reported SUI, 156 (14%) reported POP, and 251 (22%) reported AI (Table 2). Comparing the 456 women with a short IPI to the 671 women without this exposure, a short IPI was not associated with SUI (18.9% vs 19.8%, P = 0.69), POP (13.4% vs 14.2%, P = 0.71), or AI (22.4% vs 22.2%, P = 0.95). In a multivariable model controlling for age at first delivery, BMI, parity, and delivery mode, no statistically significant differences were observed between the IPI and these PFDs.

TABLE 1.

Characteristics of 1,127 Women by Presence of Short Interpregnancy Interval

	Never Had Short IPI, N=671	Ever Had Short IPI, N=456	P
Age at first delivery, y
< 30	307 (45.75)	181 (39.69)	< 0.001
30–35	272 (40.54)	155 (33.99)
≥35	92 (13.71)	120 (26.32)
Age at study entry, y
< 40	481 (71.68)	282 (61.84)	<0.001
40–45	171 (25.48)	129 (28.29)
≥45	19 (2.83)	45 (9.87)
Black race (%)	93 (13.86)	48 (10.53)	0.10
Parity at time of last survey, median (range)	2 (2–5)	2 (2–6)	< .0001
BMI at study entry
Underweight (< 18.5)	9 (1.34)	4 (0.88)	0.005
Normal (18.5-< 25)	306 (45.6)	226 (49.56)
Overweight (25-< 30)	181 (26.97)	146 (32.02)
Obese (≥30)	175 (26.08)	80 (17.54)
Education level at study entry
High school graduate or less	25 (3.73)	18 (3.95)	0.31
College degree	373 (55.59)	273 (59.87)
Graduate degree	273 (40.69)	165 (36.18)
Tobacco use (≥100 cigarettes in lifetime) at study entry	194 (29.35)	157 (34.89)	0.05
Birth types
1 = all cesarean	318 (47.39)	224 (49.12)	0.85
2 = any operative (vacuum, forceps)	88 (13.11)	57 (12.5)
3 = at least 1 vaginal birth, no operative	265 (39.49)	175 (38.38)
Macrosomia, any delivery (%)	133 (19.82)	111 (24.34)	0.07
Episiotomy, any delivery (%)	199 (29.66)	123 (26.97)	0.33
OASIs, any delivery	57 (8.49)	47 (10.31)	0.30
Prolonged second stage of labor, any delivery	159 (43.32)	108 (45)	0.68
Vacuum-assisted, any delivery	48 (7.15)	23 (5.04)	0.15
Forceps-assisted, any delivery	43 (6.41)	35 (7.68)	0.41
Longitudinal follow-up, years from first delivery to latest follow-up, median (25%, 75%)	11.2 (9.5, 13.9)	11.1 (9.0, 13.9)	0.43
Longitudinal follow-up, years from first delivery to first survey, median (25%, 75%)	6.4 (5.5, 8.1)	6.5 (5.6, 8.3)	0.22

BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); IPI, interpregnancy interval; OASI, obstetric anal sphincter injury; SD, standard deviation.

TABLE 2.

Prevalence of Pelvic Floor Disorders Based on Ever Having a Short Interpregnancy Interval and Associated Odds Ratios

Pelvic Floor Disorder	Never Had Short IPI, N=671	Ever Had Short IPI, N=456	Odds Ratio (95% CI)	Adjusted Odds Ratio (95% CI)
Stress urinary incontinence	133 (19.8)	86 (18.9)	0.91 (0.66–1.27)	1 (0.69–1.44)
Pelvic organ prolapse	95 (14.2)	61 (13.4)	1.01 (0.55–1.84)	0.83 (0.41–1.70)
Anal incontinence	149 (22.2)	102 (22.4)	1.12 (0.83–1.53)	1.08 (0.76–1.53)

Logistic regression of short IPI and various pelvic floor disorders adjusted for age at first delivery, body mass index, birth type category, and time-varying parity using a repeated measures generalized estimating equations model.

IPI, interpregnancy interval.

Two sensitivity analyses were performed: one including only the first and second delivery and a second to exclude patients with unknown gestational ages and, therefore, unknown IPI. Neither sensitivity analysis found a difference in the study outcomes. When narrowing the cohort to women with only nonoperative vaginal deliveries or spontaneous vaginal deliveries, there remained no difference in SUI, POP, or AI based on ever having a short IPI (P = 0.88, 0.84, 0.78, respectively) (Table 3). Post hoc power and sample size calculations demonstrated that, given the very small differences observed between groups, we would have required > 26,000 patients per group to detect observed differences by a short IPI in SUI, > 30,000 patients per group for POP, or > 1 million patients per group for AI.

TABLE 3.

Prevalence of Pelvic Floor Disorders in 440 Women Who Had at Least One Nonoperative Vaginal Delivery

Pelvic Floor Disorder	Never Had Short IPI, N=265	Ever Had Short IPI, N=175	P
Stress urinary incontinence	68 (25.7)	46 (26.3)	0.88
Pelvic organ prolapse	95 (14.2)	61 (13.4)	0.84
Anal incontinence	149 (22.2)	102 (22.4)	0.76

IPI, interpregnancy interval.

DISCUSSION

In our large, prospectively recruited cohort of patients derived from a single urban center in the United States, we found no association between a short IPI and the presence of SUI, AI, or POP in the first 10–15 years after first birth. We chose to study the relationship between the IPI and the development of PFDs, as a short IPI is a potentially modifiable risk factor in conception planning. ACOG has defined 18 months as the minimum recommended IPI, due to an increased risk of adverse obstetrical outcomes associated with shorter intervals. These include an increased risk of preterm birth, small-for-gestational-age fetuses, low birth weight, and uterine rupture.^22–25 Specifically for cesarean delivery, an IPI of 18 months has been shown to reduce the risk of uterine rupture, hypothesized to be due to the time required for muscle recovery and scar remodeling.²⁶ It is apparent that a short IPI can pose an immense risk in the peripartum period, but little is known about the effects of a short IPI on pelvic floor health in the decade after delivery.

Our negative results contradict the small body of research evaluating the relationship between the IPI and PFDs. A small number of other studies have considered this relationship, focusing on the interdelivery interval (defined as the interval between deliveries) rather than the IPI. For example, a study at a hospital in Shanghai, China, considered the characteristics of 2,492 postpartum women by median interdelivery interval. Urinary incontinence was assessed using the International Consultation on Incontinence Questionnaire-Urinary Incontinence Short-Form, a validated tool. They found that the highest odds for UI symptoms in the postpartum period were found among women in the lowest quartile for interdelivery interval (< 41 mo). In that study, however, the relationship between the interdelivery interval and postpartum incontinence was not linear: the odds for self-reported postpartum urinary leakage were lowest for women with a median interdelivery interval of 72 months.²⁷ In addition, this study only surveyed women up to 60 days from delivery and cannot comment on chronic, long-term incontinence that occurs or persists after the immediate postpartum recovery. Furthermore, though this research demonstrated that the risk of urinary incontinence was higher for women with an interdelivery interval of < 41 months compared with ≥ 41 months, it did not specifically address the relationship between a short interval (defined as 18 months in our present study) and UI. Thus, while there may be a broad relationship between the delivery interval and UI, it remains unclear whether this relationship exists for specifically a short IPI.

Another study of parous women in rural Nepal compared the interdelivery interval among 39 women with symptoms attributed to prolapse versus 92 controls. The cases with prolapse symptoms were more likely to report a short minimum birth interval (defined as less than 2 years), as well as older age and higher parity. Prolapse was not assessed on physical examination in this study, and the validity of the prolapse symptom assessment is uncertain.²⁸ Given that this study may not have applied a rigorous and validated assessment of PFDs to the entire population, it is unclear whether the results of this study may be biased.

A third study, conducted in a gynecologic practice in Turkey, compared 96 gynecologic patients with stage 0–1 support on POP-Q examination versus 96 patients with stage 2–4 support. In that comparison, those with stage 2–4 support reported a shorter minimum birth interval (18 vs 25 mo) and a shorter mean interval across all pregnancies (31 vs 38 mo).²⁹ Thus, these results differ from ours in that we did not identify an association between prolapse (to or beyond the hymen) and a shorter IPI. However, the prior was conducted in rural Turkey and among a population characterized by a higher mean parity (> 4), limited formal education, and low economic status, while the present study was conducted in a highly educated suburban population. It is unclear whether these distinctions contribute to the observed differences.

Studies have shown that vaginal delivery presents a greater risk to the development of PFDs compared with cesarean delivery.^10,30 Also, operative delivery, especially forceps-assisted delivery, appears to further increase the odds of PFDs.¹¹ Nonetheless, in our cohort, when we restricted the analysis to women who only had nonoperative vaginal deliveries, there was still no significant association between a short IPI and PFDs. This analysis underscores the validity of our findings across various modes of delivery.

Strengths and Limitations

The strength of this study lies in its longitudinal follow-up of the patient population. The participants in the MOAD study were recruited 5–10 years after their first birth (median: 6.4 y), then followed up for a median of 5 years afterwards. The symptoms of pelvic floor dysfunction may not become clinically significant until several years after childbirth, so it is important that our study uses data from a study population that is followed for decades.⁶ Furthermore, a strength of this study was a separate analysis of the different types of PFDs, as this allows for a more specific examination of the relationship between a short IPI and PFDs. Moreover, as has been previously described, the tools used to define PFDs have been validated.¹⁰

There are several limitations to this study as well. The MOAD study is a single-institution study, which may limit its generalizability. Furthermore, the outcome of interest in this study is the simple presence or absence of a PFD, but does not account for the severity or bother from these symptoms. It is possible that there is an association between more advanced PFDs and the IPI that are not captured due to the general prevalence of mild PFDs. In addition, this study only looked at patients in the first 10–15 years after childbirth. Almost all of the patients were younger than 45 years and, therefore, we do not know about the effect of a short IPI on PFDs in the postmenopausal years. We also do not know how many women had PFDs before their first pregnancy and delivery that were untreated and, therefore, cannot know for sure if the PFDs reported at the baseline study entry were new since birth or preceded pregnancy. In addition, the PFDs which women may experience in the first year after delivery may have resolved by the time they were recruited for the MOAD study 5 years later. Finally, as our study was underpowered to detect small effect sizes such as those we observed, the possibility of type II error (false-negative findings) cannot be ruled out.

CONCLUSION

A short IPI is not associated with PFDs in the 10–15 years following a woman’s first delivery. While a short IPI is not advised by ACOG for other reasons, our study suggests that early-onset PFDs are not related to having a short IPI. This study contributes to a growing body of research regarding the risk of PFDs due to a short IPI and prompts further study into this relationship so that obstetricians can most accurately counsel and prepare their patients.

WHY THIS MATTERS.

It is well-documented that pregnancy increases the risk of pelvic floor disorders (PFDs), which can lead to long-term effects on quality of life. While the anatomic changes of pregnancy are unavoidable, it is critical to understand how the modifiable risk factors associated with pregnancy influence the development of PFDs. One such risk factor is the interpregnancy interval (IPI). Broadly, prior studies have demonstrated a negative relationship between the IPI and the subsequent development of PFDs. However, few studies have specifically assessed a short IPI as a risk factor for PFDs. Our research, conducted via secondary analysis of the Mothers’ Outcomes After Delivery study, did not find a significant relationship between a short IPI and the presence of long-term PFDs. This finding persisted even when analyzing only vaginal deliveries, another modifiable risk factor for PFDs. This study, then, contradicts the existing body of research regarding the relationship between a short IPI and PFDs. While this difference may be attributed to population-level differences, it prompts further investigation into this subject, as understanding this relationship more can direct preconception counseling.