PHYSIOLOGY

Functional Changes

The physiology of the lower urinary tract during normal pregnancy has not been thoroughly investigated. Furthermore, significant hemodynamic changes may directly affect urinary tract function throughout the course of pregnancy. Total blood volume increases by up to 40%, and cardiac output is increased by 30% to 50% by the third trimester of pregnancy. Simultaneously, systemic vascular resistance is reduced due to progesterone-induced smooth muscle relaxation. Glomerular filtration rate and renal plasma flow consequently increase by 40% to 50%, and 60% to 80%, respectively.¹ These changes result in increased urine output and, with the pressure effect of the gravid uterus on the bladder, may cause urinary frequency, urgency, and nocturia. Up to 80% of women experience urinary frequency during pregnancy, and the symptom usually appears early (i.e., in the first trimester) and tends to worsen as pregnancy progresses.²^–⁵

Early radiologic and endoscopic studies showed that the bladder is pressed by the gravid uterus and pushed upward and anteriorly.^2,^6,⁷ More recent studies investigated the mobility of the bladder neck and pelvic floor during pregnancy. Peschers and colleagues⁸ reported that bladder neck position at rest is changed at the end of pregnancy, compared with the position in nulligravidas. This finding suggests that mechanical pressure of the fetal head and uterus and/or hormonal and connective tissue changes may alter the bladder neck position during pregnancy. Similarly, Wijma and associates⁹ used perineal ultrasound to study a cohort of 117 pregnant nulliparous and 27 nonpregnant nulliparous women. The angle of the urethrovesical junction at rest and the displacement/pressure coefficient during coughing showed a significant increase during pregnancy. King and Freeman,¹⁰ also using perineal ultrasound, found that women with stress urinary incontinence at 10 to 14 weeks postpartum had significantly greater bladder neck mobility antenatally than those who were continent postpartum. There were no significant differences in obstetric parameters between the postpartum continent and incontinent groups. The authors suggested that collagen susceptibility to changes during pregnancy, measured by changes in bladder neck mobility, could predict postpartum incontinence. Dietz and colleagues¹¹ investigated pelvic organ mobility in 28 pregnant women, seen at 10 to 17 weeks and again at 32 to 39 weeks of gestation, and 88 nonpregnant controls. Pregnant women showed greater bladder and urethral mobility. The effect was already noticeable in the first trimester, but it was significantly augmented later in pregnancy. Because similar changes were also noted in elbow hyperextension, the authors speculated that pregnancy may adversely affect connective tissue biomechanics. This speculation is further supported by earlier data suggesting that, compared with nonpregnant women, connective tissue in pregnant women contains less collagen and has less tensile strength, greater extensibility under low pressure, and loss of recoil after overstretching.¹²

Urodynamic Observations

Only a few urodynamic studies have been performed on pregnant women. Van Geelen and coworkers ¹³ assessed the urethral pressure profile during pregnancy and after delivery in 42 nulliparas. Lower values of urethral pressure profile parameters were observed in almost all women who experienced stress urinary incontinence during pregnancy and/or after delivery. The investigators concluded that “inherent weakness of the urethral sphincter mechanism plays a key role in the pathogenesis of stress incontinence.”¹³ Other urodynamic studies revealed decreased cystometric parameters during pregnancy. Chaliha and colleagues¹⁴ studied 286 nulliparous women during the third trimester of pregnancy, 161 of whom returned for postpartum urdynamic evaluation. The mean urodynamic values in the third trimester and after delivery were lower than the reported normal limits of the nonpregnant population. Antenatally, the prevalence of urodynamically proven stress incontinence was 9% and that of detrusor instability was 8%; after delivery, these values were 5% and 7%, respectively. Similarly, Nel and associates¹⁵ studied 66 pregnant women, 40 of whom returned for postpartum urodynamic evaluation. A strong desire to void, urgency, maximum cystometric capacity, and maximum and average flow rates were all significantly decreased during pregnancy. Urodynamically proven stress incontinence was present in 12% of the pregnant subjects, and detrusor instability in 23%.

The cystometric capacity of the bladder during pregnancy has been studied by several investigators, but findings are contradictory. An early study by Muellner ¹⁶ found that bladder capacity gradually increased from 12 to 32 weeks of gestation, reaching an average capacity of 1300 mL, and was associated with bladder hypotonia. This finding was confirmed by another small urodynamic study¹⁷ but doubted later by Francis,² who found a reduced bladder capacity in the third trimester. Francis² also found increased detrusor irritability in late pregnancy, rather than bladder hypotonia. More recent studies have confirmed these latter observations.^14,¹⁵

PATHOPHYSIOLOGY

Labor and delivery have long been known to be the major causes of pelvic floor injury. However, it is unknown whether the insult begins during pregnancy, before the active process of labor and delivery, and whether delivery by elective cesarean section, with no trial of labor, provides any protective effect. Major injury mechanisms include partial denervation of the pelvic floor and direct injury to the pelvic soft tissues.

Neurologic Damage

Electromyography (EMG) and pudendal nerve terminal motor latency (PNTML) measurements are considered to be useful in detecting denervation of the pelvic floor. Prolonged PNTML is obtained when large myelinated nerve axons have been damaged. Snooks and colleagues ^18,¹⁹ used electrophysiologic techniques to study 71 women at 48 to 72 hours after delivery and again, in 70% of these women, 2 months later. An increased PNTML was found in 42% of the women 48 to 72 hours after vaginal delivery, but not in any of those who delivered by cesarean section. Multiparity, forceps delivery, increased duration of the second stage of labor (defined as the interval between full cervical dilatation and the delivery of the newborn), third degree perineal tear and high birth weight were all found to be associated with increased risk of pudendal neuropathy. However, by 2 months postpartum, the PNTML had returned to normal in 60% of the women, implying that the denervation injury is usually reversible. Fourteen multiparas of the original cohort underwent repeated neurophysiologic studies 5 years after delivery.²⁰ Five of these 14 women complained of stress urinary incontinence and were found to have marked pudendal neuropathy. The investigators concluded that childbirth-associated pudendal neuropathy may persist and worsen with time.

Allen and colleagues ²¹ studied the innervation of the pelvic floor muscles before and 2 months after delivery in 96 nulliparas. Using motor unit potential duration as a sign of reinnervation in response to denervation injury, they found evidence of partial denervation of the pelvic floor with consequent reinnervation in 80% of the women after vaginal delivery. It was unclear whether the EMG changes were due to stretching of the pudendal nerve or to direct pressure of the fetal head on the nerve branches. Furthermore, evidence of partial denervation was also found in women who had undergone cesarean section during labor. This finding suggests that the denervation process may start during the first stage of labor, before delivery. Women who had a long (>56.7 min) active second stage of labor (defined as the stage of active pushing) and heavier (>3.41 kg) babies showed the most EMG evidence of nerve damage. The investigators concluded that vaginal delivery causes partial denervation of the pelvic floor in the majority of women delivering their first baby. For most women, the degree of denervation is slight, but in some there is severe damage that may be associated with loss of sphincteric control. Of the original study cohort of 96 women, 77 (80%) were available for 7-year and 65 (68%) for 15-year follow up.²² The motor unit potential durations were found to be increased significantly after delivery, and again at 7 and 15 years; however, no correlation was found between this EMG finding and the symptom of stress incontinence. The investigators concluded that the absence of an adequate marker for pelvic floor denervation makes it difficult to determine the role of denervation/reinnervation after the first delivery in the etiology of stress urinary incontinence.

Muscular Damage

Relative weakness of the pelvic floor muscles after vaginal delivery was confirmed in several clinical studies. Insult may be secondary to nerve damage, local ischemia, muscle distention, or direct tearing of muscle fibers. Peschers and associates ²³ demonstrated that pelvic muscle strength was significantly reduced 3 to 8 days after vaginal delivery but not after cesarean section. In most women, muscle strength was restored to antepartum values 6 to 10 weeks postpartum. However, Sampselle and coauthors²⁴ reported that recovery of levator ani contractility could take as long as 6 months after delivery.

Advanced imaging techniques have enabled visualization of the pelvic floor structures before and after labor and delivery. Sultan and colleagues ²⁵ used endosonography to assess antenatal and postnatal anal sphincter anatomy. At 6 to 8 weeks postpartum, 35% of the 79 primiparous women studied had occult disruption of the internal or external anal sphincter. None had such sphincter defects before delivery. Of 48 multiparous women, 40% had a sphincter defect before delivery and 44% thereafter. None of the 23 women who underwent cesarean section had a new sphincter defect after delivery. Further analysis revealed that forceps delivery was significantly associated with anatomic damage.

Magnetic resonance imaging (MRI) has been used to detect anatomic and chemical changes, as well as to localize specific injury sites. DeLancey and coworkers ²⁶ used MRI to explore the appearance of the levator ani muscle after vaginal delivery. The study population consisted of 80 nulliparous and 160 primiparous women. The primiparas were all examined 9 to 12 months after vaginal delivery. As many as 20% of the primiparas were found to have levator ani defects on MRI. Most defects were identified in the pubovisceral portion of the levator ani (consists of the pubococcygeus, puborectalis, and puboperineus muscles), and some were in the iliococcygeal portion of the muscle. No levator ani muscle defects were identified in nulliparous women. Moreover, stress-incontinent women were twice as likely to have levator ani defects than continent women. More recently, Lien and colleagues²⁷ used MRI to create a three-dimensional computer model of the levator ani muscle. This model was used to quantify levator ani muscle stretch during the second stage of labor. The investigators found that the medialmost pubococcygeus muscle is at greater risk for stretch-related injury than any other levator ani muscle during the second stage of labor. Tissue stretch ratios were also proportional to fetal head size.

Connective Tissue Damage

Normally, the endopelvic fascia and the anterior vaginal wall form a hammock-like layer in which the bladder and vesical neck rest. During increased intra-abdominal pressure, the urethra is compressed against this supporting hammock, and continence is maintained. Simultaneous contraction of the levator ani and the urethral sphincter muscles must also occur to support the vesical neck and to occlude the urethra.²⁸ Indirect evidence of the effects of childbirth on this coordinated support mechanism was obtained by sonogram measurements of the vesical neck position before and after delivery. Both transperineal and transvaginal ultrasound were used to facilitate visualization of the vesical neck at rest; then, with the Valsalva maneuver, the relative descent was measured. Several studies showed lower vesical neck position in women who delivered vaginally, compared with those who underwent elective cesarean section and with nulliparous women. Likewise, mobility of the vesical neck during the Valsalva maneuver was found to increase after vaginal delivery. However, it is less clear whether this increased vesical neck mobility is also associated with long-term pelvic floor disorders.^8,^10,²⁹^–³¹

OBSTETRIC RISK FACTORS

Parity, prolonged labor, instrument-assisted delivery, and increased birth weight have always been considered predisposing factors for pelvic floor injury and subsequent development of long-term pelvic floor dysfunction. However, clear scientific data regarding various obstetric parameters, as well as the possible protective effects of cesarean section, are inconsistent. These conflicting data stress the need for further investigation of the role of pregnancy and delivery in the development of pelvic floor disorders.

Parity

Several investigators reported a positive correlation between parity and stress urinary incontinence.³²^–³⁵ However, data concerning a possible linear correlation versus a certain threshold of parity for the development of urinary incontinence are subject to controversy. Foldspang and colleagues³⁶ found the prevalence of urinary incontinence in women aged 30 to 44 years to be correlated with parity. However, in women aged 45 years or older, only three or more deliveries were associated with an increased risk of incontinence. Thomas and associates,³⁷ in a postal questionnaire study of more than 7000 women, reported an increased prevalence of urinary incontinence in women with four or more children. Nulliparous women had a lower prevalence of urinary incontinence than did those who had been delivered of one, two, or three babies, but within the parity range of one to three there were no differences in prevalence. Similarly, Wilson and co-workers³⁸ reported that the odds ratio for postpartum urinary incontinence increased significantly after four deliveries.

The association between parity and urinary incontinence was also investigated by two large epidemiologic studies, the Norwegian Epidemiology of Incontinence in the County of NordTrondelag (EPINCONT) study ³⁹ and the Nurses’ Health Study.⁴⁰ The EPINCONT study was a large survey performed in one county in Norway during the years 1995-1997. The association between parity and urinary incontinence was analyzed in an unselected sample of 27,900 women who answered a detailed questionnaire. Overall, urinary incontinence was reported by 25% of participants. Parity was associated only with stress and mixed types of incontinence, the first delivery being the most significant. The association was strongest in the age group 20 to 34 years, with relative risks of 2.2 and 3.3 for primiparas and grand multiparas, respectively. A weaker association was found in the age group 35 to 64 years (relative risks, 1.4 and 2.0, respectively), and no association was found among women older than 65 years of age. The investigators concluded that all effects of parity seem to disappear in later years. The Nurses’ Health Study comprised 83,168 women aged 50 to 75 years. Overall, urinary incontinence was reported by 34% of participants. Similar to the EPINCONT study results, parity was associated with increased prevalence of urinary incontinence; however, the association was weaker among women aged 60 years or older. Results of these two large epidemiologic studies suggest that additional factors, other than childbirth, become more significant in older women, thereby minimizing the effects of parity per se.

Birth Weight

The importance of specific labor parameters and their etiologic role in the development of pelvic floor disorders remains controversial. Dimpfl and associates ⁴¹ found a similar incidence of postpartum stress urinary incontinence among mothers whose neonates weighed 3500 g or more and those with infants weighing less than 3500 g. Viktrup and coworkers⁴² reported increased birth weights in infants of mothers who developed stress urinary incontinence after delivery, although statistical significance was not reached. Our group³⁴ found that the prevalence of postpartum stress incontinence among grand multiparas who had given birth to at least one baby weighing more than 4000 g was significantly higher than in those who had not (29.4% versus 16.7%, respectively). Similarly, Persson and colleagues³⁵ found that the risk of later surgery for stress incontinence after vaginal delivery correlated with the weight of a woman’s largest infant.

The effect of various obstetric parameters and urinary incontinence in later life was also investigated in the Norwegian EPINCONT study.⁴³ The investigators analyzed data from 11,397 women who had delivered vaginally only and who had no more than five children. Nine obstetric parameters were investigated: birth weight, gestational age, head circumference, functional delivery disorders, injuries/tears, breech, forceps, vacuum deliveries, and epidural anesthesia. Statistically significant associations were found between any incontinence and birth weight of 4000 g or more and between stress urinary incontinence and high birth weight; however, odds ratios were relatively weak (1.1 and 1.2, respectively).