Do measures of muscular fitness modify the effect of intra-abdominal pressure on pelvic floor support in postpartum women?

Abstract

Objective:

To determine whether measures of muscular fitness modify the effect of intraabdominal pressure (IAP) during lifting on pelvic floor support.

Methods:

Participants, primiparous women 1 year following vaginal delivery, underwent the Pelvic Organ Prolapse Quantification examination, measurement of IAP via a vaginal sensor while lifting a weighted car seat, pelvic floor muscle (PFM) force assessment using an instrumented speculum, grip strength using a hand dynamometer and trunk flexor endurance (TFE) by holding an isometric contraction while maintaining a 60-degree angle to the table. We dichotomized pelvic floor support as worse (greatest descent of the anterior, posterior, or apical vagina during maximal strain at or below the hymen) vs better (all points above the hymen).

Results:

Of 825 participants eligible after delivery, 593 (71.9%) completed 1-year study visit. Mean age was 29.6 years (SD 5.0). One year postpartum, 55 (9.3%) demonstrated worse support. There were no differences in IAP during lifting or in other measures of pelvic floor loading between women with better and worse support. In multivariable analyses, neither grip strength nor PFM force modified the effect of IAP on support. In women with TFE duration ≥13 minutes, the odds of worse support increased significantly as IAP increased. No fitness measure modified the effect of other measures of pelvic floor loading on support.

Conclusions:

Primiparous women with higher IAP during lifting and greater muscular fitness did not have reduced odds of worse pelvic floor support, compared to those with lower IAP at the same muscular fitness.

ONE SENTENCE SUMMARY:

Primiparous women with higher intraabdominal pressure (IAP) during lifting and greater muscular fitness did not have reduced odds of worse pelvic floor support, compared to those with lower IAP at the same muscular fitness.

INTRODUCTION

Currently, the clinical practice of pelvic organ prolapse (POP) is focused on treatment, which generally occurs years after the primary event that initiates the onset of pelvic floor deterioration, vaginal childbirth.¹ Despite the fact that up to 1 in 5 women undergoes surgery for POP in her lifetime, we understand little about how to prevent this common disorder.^2,3 The greatest risk for POP is incurred at the time of first vaginal delivery, but it appears that a myriad of factors integrate to create a threshold condition where symptoms begin to appear.^4–6 Pregnancy and childbirth may cause deleterious changes not only to the pelvic floor, but also to the abdominal wall and overall fitness; these changes recover in many, but not all women, by one year postpartum.^7–11 A better understanding of events that impact recovery following vaginal childbirth might help us to prevent some women from crossing the threshold to POP.^12,13

One such event may be increased intraabdominal pressure (IAP), implicated as a potential causal factor in the pathogenesis of POP.¹⁴ It is possible that women who have higher levels of muscular fitness may have better pelvic support even when IAP increases during specific activities. It may be that women with greater upper arm strength perform certain activities like lifting without increasing IAP to the same degree as women less strong. Or, it is possible that women with greater trunk muscle fitness are able to achieve appropriate postural support that reduces strain on the pelvic floor muscles when IAP increases with lifting or other tasks. Exploring muscular fitness as a potential mechanism that may modify the relationship between pelvic floor support and IAP could provide a realistic target for POP prevention. This would be particularly relevant for the population of women that have recently incurred their first major insult to the pelvic floor, primiparous women delivered vaginally

Thus, the primary aim of this cross-sectional data analysis is to determine whether measures of muscular fitness, including pelvic floor muscle (PFM) force, grip strength (a good reflection of upper arm strength¹⁵), and trunk flexor endurance (TFE) modify the effect of IAP during lifting on pelvic floor support at 1 year postpartum. The secondary aims are to determine whether these same measures of muscular fitness modify the effects of self-reported heavy lifting or work, chronic cough and constipation on pelvic floor support 1 year postpartum.

We hypothesized that women with high IAP 1 year postpartum who also demonstrate lower PFM strength, grip strength, or TFE 1 year postpartum will have higher odds of worse pelvic floor support at 1 year postpartum than women with low IAP at the same muscular fitness, whereas women demonstrating greater fitness will have more similar odds of worse pelvic floor support regardless of IAP. We hypothesized the same direction of effect for heavy work, chronic cough and constipation as for high IAP.

METHODS

This is a planned primary analysis from the Motherhood And Pelvic health (MAP) study; methods for the overarching study have been published.¹⁶ Participants in this current cross-sectional analysis completed 1-year follow-up in MAP, a prospective cohort study. We modeled the analysis in this report on the Integrated Lifespan Model described by Delancey et al.⁴ This model describes factors that might influence pelvic floor function in 3 major life phases: 1) Predisposing factors that affect the functional reserve that is established during a woman’s growth; 2) Inciting factors, primarily injury and potential recovery that occurs during and after vaginal birth; and 3) Intervening factors that affect the deterioration that occurs with age. We considered predisposing factors including ethnicity, age, family history of POP or urinary incontinence (UI) and third trimester pelvic floor support. All women underwent the inciting factor of vaginal childbirth and we also considered the additional impact of a delivery notable for the presence of a factor previously reported to increase the risk of levator ani muscle (LAM) injury.¹⁷ We considered intervening factors one year postpartum that may affect deterioration in ways other than aging, potentially by exerting higher forces on the pelvic floor (IAP during lifting, heavy work, chronic cough and chronic constipation), as well as intervening factors that might provide counterforce to the pelvic floor (TFE, grip strength, PFM force). The schematic model upon which our analysis is based is shown in Supplementary Figure 1.

This study was approved by the Institutional Review Boards at the participating sites. All participants completed written informed consent.

We recruited participants during prenatal visits at one of seven clinical sites. Participants were eligible if they were ≥ 18 years of age, nulliparous at ≥28 weeks gestation with a singleton pregnancy, were not planning to deliver by cesarean, were English- or Spanish-speaking, did not have medical problems that precluded physical activity, lived within 60 miles of the research center, and did not plan to move out of the area before 1 year postpartum. After delivery, women that delivered by cesarean or before 37 weeks gestation were excluded from further participation. Participants completed questionnaires and physical examinations during the third trimester, at 5–10 weeks postpartum and at 1 year (range 11–15 months) postpartum. For the current analysis, we excluded women that completed only the 1-year questionnaire and not the 1-year in-person visit. Procedures relevant to the current analysis are described in greater detail, as follows.

The primary outcome measure, pelvic floor support, was based on the Pelvic Organ Prolapse Quantification (POP-Q) system¹⁸, a reproducible method for assessing vaginal descent.^19,20 With participants in the low lithotomy position and with back elevated to 45 degrees, trained and certified study coordinators measured each POP-Q point to the nearest 1 cm. We dichotomized pelvic floor support using maximum vaginal descent (MVD), that is, the greatest observed descent of the anterior, posterior, or apical vagina during maximal strain, as above the hymen (< 0 cm; better support) versus at or below the hymen (≥ 0 cm; worse support). This cut-point is commonly used in research because it represents the level at which more women become symptomatic.^21,22

We assessed IAP using a vaginal sensor system developed and validated by our group.^23,24 The clinical coordinator inserted the sensor into the upper third of the vagina and with the sensor in place, the participant performed three repetitions of lifting a car seat which weighed 12.5 kg: 6.8 kg based on weight of top selling car seats at the time of study planning plus a 5.7 kg added “infant” weight, based on the 50^th percentile for weight in boys at 10 weeks in the U.S.²⁵ We chose to study lifting a weighted car seat as this is an activity done regularly by most postpartum women. We reported the mean maximal IAP as the average of the five highest peaks recorded during the lifting task that were at least 1 second apart.²⁶

To determine trunk flexor endurance, measured in seconds using a stopwatch, we used the protocol described by McGill et al. While seated on an exam table with hips and knees flexed to 90 degrees and feet secured, the participant held an isometric contraction for as long as possible in order to maintain a 60-degree angle with reference to the table as guided by a wedge positioning device.²⁷

We measured handgrip strength using a dynamometer (Model J00105, Lafayette Instruments).²⁸ The participant stood with her elbow flexed at 90 degrees, and with her dominant hand performed three maximal squeezes, each for 3–5 seconds followed by 30 seconds of rest. We recorded the highest of the three trials in kilograms (kg) of force.

We measured PFM force using an instrumented speculum designed to minimize the effect of IAP on the outcome.²⁹ As previously noted, we initially used the modified speculum described by the original investigators.³⁰ However, this was not designed to withstand the rigors of sterilization required of our study protocol, and ultimately failed to accurately measure PFM force. Therefore, we developed a similar system that could undergo repeated sterilization.³¹ While this was being developed, we were unable to conduct PFM force testing. Therefore, 138 of 593 (23.3%) women that completed the 1-year visit had no PFM force results and were excluded from this portion of the current analysis. Consistent with previous work, we defined maximal PFM force (also known as vaginal closure force) as the average of 3 maximal contractions, in Newtons (N), minus the baseline force measured in air.^29,32,33 We measured weight and height without shoes using a medical scale and wall stadiometer.

Participants completed questionnaires online in either English or Spanish. We assessed chronic cough as per the National Health And Nutrition Survey: “Do you usually cough on most days for 3 consecutive months or more during the year?” We defined constipation according to the Defecation Distress Inventory^34,35 in which constipation is present with a positive response to both of the following questions: “Thinking back on the last 4 weeks, do you have less than 3 bowel movements per week? Do you have to strain more than 25% of the time to have a bowel movement?” We assessed heavy lifting or work using activities from a questionnaire previously validated for assessing skeletal loading³⁶ and by modifying to include work-related tasks with the following question: “In the last 7 days, did you do any activity for at least 10 minutes that involves heavy lifting or work, such as lifting heavy objects (not related to caring for your baby), heavy gardening, lifting weights at the gym, or lifting, pushing, or pulling heavy objects at work?”

Study coordinators extracted delivery information from the medical record. Consistent with others, we considered a woman to be at higher risk for a LAM injury if one of the following risk factors were present: second stage of labor duration ≥ 150 minutes, birthweight > 4000 grams, forceps delivery or anal sphincter tear (3^rd or 4^th degree laceration).³⁷

In univariate analyses, we used chi-square, t-tests and logistic regression, as appropriate, to compare participant characteristics between those who did and did not have worse pelvic floor support, those that did and did not have available PFM force data, and those that completed 1-year follow-up and those that did not. We used logistic regression models that included statistical interaction terms between IAP during lifting and each measure of muscular fitness, adjusted for the predisposing and inciting factors identified in our conceptual model (Supplementary Fig. 1): ethnicity, age, high-risk delivery variable, family history of POP or UI, and third trimester pelvic floor support. As we aimed to evaluate the total effect of IAP, we did not adjust for factors that contribute to IAP, such as obesity. We derived odds ratio (OR) of IAP during lifting on pelvic floor support at various different values of muscular fitness, and used graphical displays to determine at what range(s) of muscular fitness these ORs changed from non-significant to significant. We used similar techniques to analyze the interactions between pelvic floor loading factors and each measure of muscular fitness. We used SAS version 9.4 (Cary, North Carolina) for all analyses. In our a priori sample size calculation, we considered women at the mean for IAP as “low risk” and those at the mean plus one standard deviation as “high risk”. We assumed based on available data, a frequency of worse support of 15% in the low-risk group¹⁶. Assuming that R² of IAP regressed on other predictors was 0.5, we initially planned for a sample size of 585 women which would provide 90% power to detect a minimal odds ratio of 1.82 at a significance level of 0.017 (to accommodate 3 separate binomial regression models for statistical interaction terms between each test of muscular fitness and IAP). We used this most conservative Bonferroni adjustment in the sample size calculation to ensure that the family-wise error rate was less than the conventional significance level of 0.05.

RESULTS

Of 825 participants eligible after delivery, 593 (71.9%) completed the 1-year study visit (Figure 1). The mean age of the population was 29.6 years (SD 5.0). Characteristics of the study population are summarized in Table 1. With the exception of age, there were no statistically significant differences between women with (N=455) and without (N=138) available PFM force data in any of the characteristics in Table 1. Women with available PFM force data were older, 29.8 (5.1) years vs 28.8 (4.7) years for those without PFM force data, p=0.03. Compared to enrollees that completed the 1-year visit and make up our study sample, enrollees that did not complete the 1-year visit were younger (mean 25.7 (SD 5.3) vs 28.4 (SD 5.0) years, p< 0.01), and more likely to report Hispanic ethnicity (30.2% vs 17.4%, p<0.01); there were no differences in support at or below the hymen in the third trimester or family history of POP or UI between these groups.

FIGURE 1.

Participant flow.

Table 1.

Participant Characteristics (N=593)

Characteristic*	Mean (SD) or N (%)
Age at 1 year postpartum, years; mean (SD)	29.6 (5.0)
18 to </= 25	135 (22.8%)
26 to < 33	281 (47.5%)
≥33	176 (29.7%)
Months since delivery to study visit; mean (SD)	12.6 (1.1)
Ethnicity
Non-Hispanic	490 (82.6%)
Hispanic	103 (17.4%)
Race
Caucasian/White	545 (93.5%)
Black/African American	7 (1.2%)
Asian	27 (4.6%)
Other/do not wish to identify	4 (0.7%)
Education
High school or less	61 (10.3%)
Some college/completed college	352 (59.6%)
Graduate or professional degree	178 (31.1%)
Family history of:
Pelvic organ prolapse (POP)	60 (10.7%)
Urinary incontinence (UI)	198 (35.6%)
POP or UI	208 (37.4%)
Body mass index at 1 year postpartum, kg/m2; mean (SD)	24.9 (5.7)
Normal/underweight (< 25)	352 (59.4%)
Overweight (25 to < 30)	147 (24.8%)
Obese (≥ 30)	94 (15.9%)
Work status at 1 year postpartum
Working full-time (≥30 hours per week)	303 (51.2%)
Working part-time (< 30 hours per week)	123 (20.8%)
Other (student, homemaker, disabled, unemployed)	166 (28.0%)
Breastfeeding at 1 year postpartum	269 (45.4%)
Self-assessment of physical activity level at 1 year postpartum**
Sedentary	14 (2.4%)
Underactive	381 (64.6%)
Active	195 (33.1%)
Maximal vaginal descent at third trimester
<0 cm	574 (96.8%)
≥ 0 cm	19 (3.2%)
Maximal vaginal descent at 1 year
<0 cm	538 (90.7%)
≥ 0 cm	55 (9.3%)
INTRAABDOMINAL PRESSURE (IAP), LOADING, AND STRENGTH MEASURES 1 YEAR POSTPARTUM
Maximal IAP during lifting (cmH20), mean (SD)	43.9 (12.3)
Heavy lifting or heavy work	257 (43.6%)
Constipation	22 (3.7%)
Chronic cough	12 (2.0%)
Pelvic floor muscle force (N), mean (SD)	5.2 (3.8)
Grip strength (kg), mean (SD)	32.2 (6.2)
Trunk flexor endurance duration (seconds), median (IQR)	144.5 (80.0, 248.5)
CHARACTERISTICS AT DELIVERY
Mode of delivery
Normal spontaneous vaginal delivery
Vacuum assisted delivery	13 (2.2%)
Forceps assisted delivery	45 (7.6%)
Anal sphincter laceration	23 (3.9%)
Duration of 2nd stage of labor, minutes; median (IQR)	83 (47, 142)
0–60 minutes	163 (37.2%)
61–149 minutes	179 (40.9%)
≥ 150 minutes	96 (21.9%)
Birth weight, grams; mean (SD)	3324 (397)
< 2500 grams	12 (2.1%)
2500–3000 grams	110 (18.9%)
3001– 3500 grams	277 (47.7%)
3501–4000 grams	156 (26.9%)
>4000 grams	26 (4.5%)
Presence of high-risk delivery variable	159 (26.8%)

^*The denominators do not always add up to 593 because of missing data. Percentages reflect the denominators with complete data for each variable. Family history was added to the initial questionnaire after enrollment began, accounting for a larger number of missing responses. One of the participating hospitals did not record duration of 2^nd stage of labor, again accounting for larger number of missing responses.

Table 2 summarizes the univariate analyses of potential risk factors for pelvic floor support. Women with worse support (MVD ≥ 0 cm) were older and more likely to have demonstrated worse support in third trimester. There were no differences in maximal IAP during lifting or in the proportion that reported heavy work, chronic cough or constipation between women with better and worse support. Compared to women in the lowest quartile for TFE duration, those in the upper quartile were more likely to demonstrate worse support on univariate analysis (OR 2.28 (95% CI 1.07, 4.86)). There were no differences in odds of worse support according to quartiles of grip strength or PFM force.

Table 2.

Univariate analysis of lifespan factors according to pelvic floor support

Factor	MVD <0 N= 538	MVD >/= 0 N=55	OR (95% CI)	P
PREDISPOSING FACTORS
Ethnicity				0.56
Hispanic	95 (17.66%)	8(14.55%)	REF
Non-Hispanic	443 (82.34%)	47(85.45%)	1.26 (0.58,2.75)
Age at 1 year visit, years, mean (SD) Categorical:	29.39 (4.95)	31.24 (5.32)	1.71 (1.02, 1.14)	0.01
18 to </= 25	125(23.28%)	10(18.18%)	REF
26 to < 33	264 (49.16%)	17(30.91%)	0.81 (0.36, 1.81)	0.60
≥33	148 (27.56%)	28 (50.91%)	2.37 (1.11, 5.06)	0.03
Antenatal maximal vaginal descent				0.0002
<0	526 (97.77%)	48(87.27%)	REF
>/= 0	12 (2.23%)	7(12.73%)	6.39 (2.4, 17.00)
Family history of pelvic organ prolapse or urinary incontinence				0.12
Absent	320(63.62%)	28(52.83%)	REF
Present	183(36.38%)	25(47.17)	1.56 (0.88, 2.76)
INCITING FACTOR
Presence of high-risk delivery variable				0.23
Absent	390 (72.49%)	44 (80%)	REF
Present	149 (27.51%)	11 (20%)	0.66 (0.33,1.31)
INTERVENING FACTORS
FORCE
Intraabdominal pressure during lifting, cm H2O
Mean (SD)	44.00 (12.38)	42.41 (10.98)		0.36
Quartiles
Quartile 1, 0 to 36.31	128 (24.52%)	16 (29.09%)	REF
Quartile 2, 36.32 to 42.66	131 (25.10%)	13 (23.64%)	0.79 (0.37, 1.72)	0.56
Quartile 3, 42.67 to 50.55	130 (24.90%)	15 (27.27%)	0.92 (0.44, 1.95)	0.83
Quartile 4, 50.56 to 107.45	133 (25.48%)	11 (20.00%)	0.66 (0.30, 1.48)	0.31
Heavy work				0.99
No	302 (56.45%)	31 (56.36%)	REF
Yes	233 (43.55%)	24 (43.64%)	1.00 (0.57, 1.76)
Chronic cough				0.91
No	525 (97.95%)	54 (98.18%)	REF
Yes	11 (2.05%)	1 (1.82%)	0.88 (0.11, 6.98)
Constipation				0.45
No	513 (96.07%)	53 (98.15%)	REF
Yes	21 (3.93%)	1 (1.85%)	0.46 (0.06, 3.50)
Body mass index, kg/m2; mean (SD)	25.02 (5.84)	24.06 (4.18)	0.97 (0.92, 1.02)	0.24
Normal/underweight(<25)	316 (58.74%)	36 (65.45%)	REF
Overweight (25 to < 30)	134 (24.91%)	13 (23.64%)	0.85 (0.44, 1.66)	0.63
Obese (≥ 30)	88 (16.36%)	6 (10.91%)	0.60 (0.24, 1.47)	0.26
Self-assessment of physical activity*				0.58
Sedentary or underactive	360 (67.29%)	35 (63.64%)	REF
Active	175 (32.71%)	20 (36.36%)	1.18 (0.66, 2.10)
COUNTERFORCE
Pelvic floor muscle force, N
Mean (SD)	5.32 (3.63)	5.73 (4.48)	1.03 (0.95, 1.11)	0.49
Quartiles
Quartile 1, 0.02 t 2.60	104 (25.30%)	9 (20.45%)	REF
Quartile 2, 2.61 to 4.45	100 (24.33%)	14 (31.82%)	1.62 (0.67, 3.91)	0.28
Quartile 3, 4.46 to 7.05	103 (25.06%)	11 (25.00%)	1.23 (0.49, 3.10)	0.65
Quartile 4, 7.06 to 23.98	104 (25.30%)	10 (22.73%)	1.11 (0.43, 2.85)	0.83
Trunk flexor endurance, sec
Mean (SD)	189.03(174.10)	304.35 (303.64)	1.002 (1.001, 1.003)	0.0001
Quartiles
Quartile 1, 0 to 79	135 (25.33%)	11 (20.00%)	REF
Quartile 2, 80 to 145	140 (26.27%)	8 (14.55%)	0.70 (0.27, 1.80)	0.46
Quartile 3, 146 to 249	134 (25.14%)	13 (23.64%)	1.19 (0.52, 2.75)	0.68
Quartile 4, 250 to 1502	124 (23.26%)	23 (41.82%)	2.28 (1.07, 4.86)	0.03
Grip strength (N)
Mean (SD)	32.06 (5.98)	33.29 (7.20)		0.17
Quartiles
Quartile 1, 10.00 to 27.90	119 (22.20%)	10 (18.18%)	REF
Quartile 2, 28.00 to 31.90	147 (27.43%)	13 (23.64%)	1.05 (0.45, 2.48)	0.91
Quartile 3, 32.00 to 36.11	137 (25.56%)	16 (29.09%)	1.39 (0.61, 3.18)	0.44
Quartile 4, 36.12 to 56	133 (24.81%)	16 (29.09%)	1.43 (0.63, 3.28)	0.40

In separate multivariable analyses (Table 3), each containing the relevant interaction terms of maximal IAP during lifting with each of the muscular fitness measures, TFE duration was the only fitness measure that modified the effect of IAP during lifting on support (p-value for the interaction term between IAP during lifting and TFE duration was 0.03). The interaction effects between each strength measure and IAP during lifting are demonstrated in Figure 2. These figures demonstrate that there is no pattern of any modified effect between grip strength and IAP or between PFM force and IAP. At a lower TFE, the effect of IAP on worse support was not statistically significant (e.g. at TFE=3 min, OR of 1 SD increase on IAP was 0.92 (95% CI 0.66, 1.28); when TFE duration was very long, 13 minutes or more, the odds of worse support increased significantly in women with higher IAP (e.g. at TFE=15 min, OR of 1 SD increase on IAP was 3.58 (95% CI 1.03, 12.48).

Table 3.

Individual multivariable analyses testing interactions between intraabdominal pressure (IAP) and each fitness measure

	Pelvic floor muscle (PFM) force	Trunk flexor endurance (TFE)	Grip strength	PFM Force, TFE and Grip strength
Number of observations in each model	423	541	539	421
Overall Likelihood Ratio Test	p=0.046	p < 0.0001	p = 0.002	p=0.001
FACTORS	P value OR (95% CI)	P value OR (95% CI)	P value OR (95% CI)	P value OR (95% CI)
PREDISPOSING FACTORS
Ethnicity	p=0.26	p=0.85	p=0.63	p=0.47
Hispanic (REF)
Non-Hispanic	1.83 (0.64, 5.20)	1.09 (0.44, 2.73)	1.26 (0.50, 3.16)	1.49 (0.51, 4.37)
Age at 1 year visit, years	Overall p=0.01	Overall p=0.03	Overall p=0.02	Overall p=0.01
18 to </= 25 (REF)
26 to < 33	0.55 (0.21, 1.42)	0.63 (0.26, 1.53)	0.74 (0.31, 1.76)	0.42 (0.16, 1.15)
≥33	1.69 (0.69, 4.10)	1.61 (0.68, 3.78)	1.84 (0.80, 4.24)	1.41 (0.56, 3.56)
Antenatal maximal vaginal descent	p=0.056	p=0.001	p=0.0004	p=0.10
<0 (REF)
>/= 0	3.51 (0.97, 12.77)	6.32 (2.09, 19.15)	6.77 (2.33, 19.68)	3.16 (0.81, 12.33)
Family history of pelvic organ prolapse or urinary incontinence	p=0.88	p=0.08	p=0.09	p=0.99
Absent (REF)
Present	1.05 (0.54, 2.06)	1.73 (0.93, 3.21)	1.69 (0.93, 3.08)	1.00 (0.50, 2.04)
INCITING FACTOR
Presence of high-risk delivery variable	p=0.50	p= 0.36	p=0.35	p=0.50
Absent (REF)
Present	0.76 (0.34, 1.69)	0.72 (0.35, 1.48)	0.71 (0.35, 1.45)	0.76 (0.33, 1.71)
INTERVENING FACTORS
IAP during lifting, per 10 cm H2O	p=0.77 1.08 (0.66, 1.76)	p=0.09 0.71 (0.47, 1.06)	p=0.80 0.83 (0.20, 3.47)	p=0.98 0.97 (0.16, 6.02)
PFM Force	p=0.27 1.22 (0.86, 1.73)	Not included	Not included	p=0.27 1.23 (0.85, 1.76)
IAP*PFM Force	p=0.36 0.96 (0.89, 1.04)	Not included	Not included	p=0.24 0.94 (0.88, 1.03)
TFE, per 1 minute	Not included	p=0.17 0.78 (0.55, 1.11)	Not included	p=0.05 0.63 (0.40, 1.01)
IAP*TFE	Not included	p=0.03 0.10 (1.01, 1.20)	Not included	p=0.01 1.16 (1.04, 1.30)
Grip strength	Not included	Not included	p=0.89 1.01 (0.83, 1.23)	p=0.72 1.05 (0.82, 1.33)
IAP*Grip strength	Not included	Not included	p=0.90 1.003 (0.96, 1.05)	p=0.74 0.99 (0.94, 1.05)

FIGURE 2. Interaction effect between intraabdominal pressure during lifting and 3 fitness measures.

The solid lines represent the predicted odds ratios for worse support according to each of the 3 strength measures between women whose intraabdominal pressure (IAP) during lifting a weighted car seat is at the mean plus 1 standard deviation (SD) vs those at the mean (blue line) and at the mean minus 1 SD vs those at the mean (red line). The shaded areas represent the 95% confidence intervals. In Figure 2A, the white area beginning at 13 minutes, in which the two confidence interval bands do not overlap, represents a statistically significant modified effect of IAP by trunk flexor endurance on worse support. In Figure 2B, the 2 confidence interval bands overlap for the entire range of pelvic floor muscle force, demonstrating no statistically significant interaction between IAP and pelvic floor muscle force on worse support. Similarly, in Figure 2C, the confidence interval bands overlap for the entire range of grip strength, demonstrating no statistically significant interaction between IAP and grip strength on worse support.

In separate multivariable models testing secondary predictors, with one exception, the three fitness measures did not modify the effect of heavy work, chronic cough, or constipation on support (data not shown). TFE modified the effect of heavy work (p value for interaction term 0.03) on worse support. Similar to the primary results, for women that reported heavy work, a longer TFE duration increased the odds of worse support (OR 1.24 (95% CI 1.10, 1.39)) while no significant effect was seen in women that did not report heavy work (OR 1.04 (95% CI 0.94, 1.16)). TFE did not modify the effect of cough (p value for interaction 0.76) or constipation (p value for interaction 0.63) on support.

The correlations between the three fitness measures were weak with Pearson’s correlation coefficient r values of 0.14 between grip strength and TFE, 0.06 between grip strength and PFM force, and 0.14 between TFE and PFM force.

DISCUSSION

We found no association between measures reflecting forces on the pelvic floor, including IAP during lifting, and self-reported heavy lifting or work, chronic cough or constipation, on pelvic floor support in women 1 year following vaginal childbirth. In terms of counterforces (muscular fitness), on univariate analyses, women in the highest quartile of TFE were more likely to demonstrate worse support but there were no associations between grip strength or PFM force and support. Only one measure of muscular fitness, TFE, modified the effect of IAP on support, but in the opposite direction we had originally hypothesized: women with higher IAP and longer TFE duration had higher odds of worse support compared to women with lower IAP and longer TFE duration, while there was no difference in odds of worse support between women with shorter TFE duration according to IAP group. However, as seen in Figure 2, the confidence intervals are very wide, given that only 18 women demonstrated a TFE duration of 13 or more minutes.

We had hypothesized that higher IAP would be associated with worse support and that greater muscular fitness would mitigate the effect of higher IAP on pelvic floor support. However, final models not only found no association between pelvic floor support and IAP and other measures of pelvic floor loading, but also found no protective effect of greater muscular fitness. Instead, the effect of high IAP during lifting on support was greater in women with higher TFE duration. Co-contraction of trunk musculature is tightly linked to the extent of IAP generation during isometric trunk flexion and extension.³⁸ It may be that women with exceptional trunk muscle endurance, which ostensibly signifies stiffer muscles and a high level of core fitness, generate greater chronic forces on the pelvic floor, perhaps by their ability to volitionally generate higher IAP during strenuous activities. We previously found no significant relationship between TFE duration and IAP during lifting at 5–10 weeks postpartum³⁹, or between TFE duration and IAP during the TFE task at 1 year postpartum⁴⁰ but the cumulative effect of greater TFE might not be identified by measuring maximal IAP during a given activity. In our population, the sample of women for whom prolonged TFE duration incurred greater risk for worse support was small. It is possible that the interaction identified between TFE duration and IAP during lifting may be due to unexplained confounding. Future research focusing specifically on postpartum women at upper extremes of fitness would be useful to tease out the epidemiological and biomechanical factors that might play a role in the association we identified.

We are not aware of other studies that have investigated associations between measures of muscular fitness and pelvic organ prolapse. In women age 70 years and greater, there was no association between grip strength and the prevalence of monthly SUI, but SUI prevalence was greater in women with greater quadriceps torque and was not affected by walking speed. In the same study, over a 3-year period, decreased grip strength was associated with greater odds of new or persistent SUI, though the converse was not true.⁴¹ In our population, we previously found no correlation between PFM force and either grip strength or TFE 1 year postpartum.³⁰ In this current investigation, PFM force was not associated with vaginal support.

Strengths of our study include the large sample size, objective assessment of pelvic floor support, and assessment of various measures of pelvic floor force and potential counterforce. We did not standardize the way in which women lifted the car seat, in order to approximate the way in which women lift in general and reflect the true variability in this task; conversely, we did standardize the way in which women performed the 3 muscular fitness measures. As one composite measure of muscular fitness does not exist, we chose measures that are hierarchically associated with pelvic floor support, assessing PFM, trunk and upper body strength. However, it may be that other measures of fitness have different effects on pelvic floor support. Our results apply only to IAP generated by lifting a 12.5 kg car seat. We don’t know whether the same conclusions would apply to an activity that increases IAP with acceleration forces as in jumping, or the cumulative effects of chronically elevated IAP loading of the pelvic floor.

Our study was powered a priori to test the interaction terms of counterforces with IAP during lifting. In our sample size calculation, we assumed that most women would complete these 3 fitness measures but because of the technical problems described, our population for the analyses with PFM force were about one-fifth less than anticipated. The number of observations was substantially smaller for the two models containing PFM force because of the larger amount of missing data for this variable. Thus, these models also have less power than do the others, as reflected by differences in p values and odds ratios for some variables, compared to the other models. In addition to objectively measured IAP during a common lifting task, we evaluated 3 self-reported measures of pelvic floor loading (chronic cough, constipation and heavy work or lifting) that are widely considered to be potential risk factors for pelvic organ prolapse.⁴² We defined chronic cough according to the definition used by HANES, but as others have pointed out, existing epidemiologic definitions lack clinical validation.⁴³ Cough itself is quite commonly reported, by 31% of women in a population based sample,⁴⁴ and as we were more interested in chronic loading events, we restricted the definition to chronic cough with a resulting small sample size. Similarly, we also used a more stringent definition for constipation, requiring the presence of both infrequent bowel movements and straining. The prevalence of constipation in our population, 3.7 %, is thus at the lower end of that reported in adults in one systematic review.⁴⁵ Our analyses of secondary aims evaluating both of these risk factors are thus underpowered. As noted, because obesity increases IAP and we were interested in studying the total effect of IAP rather than factors that are associated with higher IAP, we decided a priori not to adjust for obesity in our models.

Our results are specific to largely healthy Caucasian primiparas that delivered vaginally and cannot be generalized to multiparas or to all racial groups. Our final study sample was not entirely reflective of the population eligible after delivery, as women that completed the 1-year study visit were older and less likely to be Hispanic than those that did not.

In conclusion, contrary to popular opinion and to our hypothesis, we did not find that women with higher IAP during lifting but greater muscular fitness had reduced odds of worse pelvic floor support, compared to women with higher IAP and lower muscular fitness. Further investigation is needed to determine whether these results are limited to postpartum women or if they are consistent in older women or in heterogeneous populations. Muscular fitness is a central component to published physical activity guidelines, and some reports specifically document health benefits of core muscle fitness.^46–48 However, the role of trunk muscle fitness, relative to IAP and pelvic floor support in postpartum women, warrants further investigation perhaps with randomized intervention and a long follow-up period.

Supplementary Material

Supplemental Data File

SUPPLEMENTARY FIGURE 1. Schematic model for current analysis, based on the Integrated Lifespan Model.