Secondary Logo

Share this article on:

Antenatal Prediction of Postpartum Urinary and Fecal Incontinence



Objective To investigate the effect of pregnancy and delivery on continence and to assess whether physical markers of collagen weakness can predict postpartum urinary and fecal incontinence (including incontinence of flatus).

Methods In a prospective, longitudinal study in a London teaching hospital, 549 nulliparas were interviewed after 34 weeks' gestation and again 3 months postpartum regarding urinary and fecal symptoms before and during pregnancy and after delivery. Family histories of incontinence, prolapse, and collagen abnormalities were recorded also. Physical examination was done to assess markers of collagen weakness such as striae, hernia, varicose veins, and joint mobility.

Results The prevalence of urinary incontinence before, during, and after pregnancy was 3.6%, 43.7%, and 14.6%, and rates of fecal incontinence were 0.7%, 6.0%, and 5.5%, respectively. Fecal urgency was more common in women who had spontaneous and instrument-assisted vaginal deliveries (n = 413) compared with cesareans (n = 131) (7.3% versus 3.1%; P = .046). Postnatal urinary or fecal dysfunction was not related to antenatal body mass index, smoking, race, striae, varicose veins, hemorrhoids, or family history of incontinence. Higher joint-mobility scores were associated with incontinence of flatus (P = .021) but not with other urinary or fecal symptoms.

Conclusion Although collagen weakness was previously implicated in the pathogenesis of incontinence, physical markers in this study could not predict postpartum urinary and fecal incontinence. Either those markers were not representative of collagen weakness, or a larger study with longer follow-up is necessary.

Clinical markers of collagen weakness such as striae, hernia, joint hypermobility, and varicose veins do not predict postpartum incontinence.

Urogynecology Unit, St. George's Hospital, London; Princess Anne Hospital, Southampton; and Mayday University Hospital, Surrey, United Kingdom.

Address reprint requests to: Charlotte Chaliha, MA, MBBChir, Urogynecology Unit, Department of Obstetrics and Gynecology, St. George's Hospital, Cranmer Terrace, Blackshaw Road, London SW17 0RE, United Kingdom

Supported by the South Thames Project Grant Scheme.

The authors thank Professor Martin Bland for advice regarding statistical analysis of the data.

Received August 27, 1998. Received in revised form April 2, 1999. Accepted April 8, 1999.

Urinary and fecal incontinence can have a devastating effect on a woman's quality of life. Vaginal delivery has been implicated as a predisposing factor, and this might increase demand for cesarean deliveries to protect pelvic-floor function. The high prevalence of postpartum symptoms independent of parity suggests that there might be an individual predisposition to pelvic-floor weakness and consequent incontinence. Although collagen abnormalities have been associated with increased joint mobility, prolapse, and urinary incontinence,1–6 their relation to fecal incontinence is not fully understood.7,8

Increased joint mobility is believed to indicate systemic alterations in collagen during pregnancy, as in women with varicose veins, and is associated with stretching of the pelvic ligaments during pregnancy.9 If collagen abnormalities predispose to incontinence, they might be inherited defects detectable by family histories. Identification of obstetric variables that can increase the risk of pelvic-floor trauma in women with inherent risks of incontinence might enable modification of obstetric practice and appropriate counseling. The aim of this study was to evaluate the use of physical markers of collagen weakness in predicting postpartum fecal and urinary incontinence and to identify associated obstetric risk factors.

Back to Top | Article Outline


We interviewed consecutive women who were nulliparous, English-speaking, and had singleton pregnancies in the third trimester. A structured questionnaire was completed by each woman for this pilot study, and each signed an informed consent. Exclusion criteria included active urinary tract infection, history of urinary tract abnormality, recurrent urinary tract infection, history of anorectal surgery or trauma, and active anorectal infection. The women were interviewed during their routine antenatal visits in clinics in which the interviewer (CC) was present between April 1996 and August 1997. At initial interviews, personal and family histories of incontinence, prolapse, and collagen abnormalities were recorded. The urinary symptom questionnaire asked about frequency, nocturia, urgency, and stress and urge incontinence. Frequency was defined as voiding seven or more times per day or more than once every 2 hours. Nocturia was defined as voiding on two or more occasions during the night. Stress incontinence was defined as loss of urine on physical effort or coughing, and urge incontinence as loss of urine associated with a strong desire to void. The fecal symptom questionnaire asked about fecal urgency and incontinence. Fecal urgency was defined as the inability to defer defecation for longer than 5 minutes.10 Fecal incontinence included incontinence of flatus and liquid and solid stool and soiling of underwear. Severity of incontinence was assessed by the degree and frequency of incontinence, pad usage, and the effect on daily activities.

Physical examination was done to assess physical markers of collagen weakness such as striae, hernia, joint hypermobility, and varicose veins. Joint-mobility scores were assessed using modified Beighton criteria.11 The women were interviewed 3 months postpartum in person or by telephone, and the same symptom questionnaire was completed. Obstetric variables evaluated were mode of delivery, epidural usage, augmentation of labor, duration of labor, perineal trauma, fetal weight, and fetal head circumference. These data were collected from patient records after the postpartum interview.

The data were computerized and analyzed using the Statistical Package for Social Sciences (SPSS Inc., Chicago, IL). Associations between categoric variables were assessed using the χ2 test, Fisher exact test, and McNe-mar test for matched data. Forward, stepwise, logistic regression analysis was used to determine whether obstetric variables were related to postpartum symptoms. P < .05 was considered statistically significant. All results are reported as mean ± standard deviation (SD). For common risk factors, we had better than 80% power to detect relative risks of 2 or more.

Back to Top | Article Outline


Five hundred forty-nine women were interviewed antenatally and all completed postpartum follow-up. The ethnic distribution was 382 white (69.6%), 82 black (14.9%), 84 Asian (15.3%), and one Southeast Asian (0.2%), which was similar to that of the institution. The obstetric and neonatal data of the study population were compared with those of the overall nulliparous, registered population delivered over the same period, and there were no statistically significant differences (Table 1). The mean gestation at interview was 37 weeks (range 34–42) and the mean age at interview was 29 years (range 17–46). In this study, 133 women (24.2%) had cesareans: 46 (35%) before the onset of labor, 79 (59%) during the first stage, and eight (6%) during the second stage.

Table 1

Table 1

The mean duration of labor was 591 ± 363 minutes for the first stage, 37 ± 33 minutes for the passive second stage (defined as time from full cervical dilatation to the onset of maternal pushing), and 55 ± 37 minutes for the active second stage (defined as time from the onset of maternal pushing to delivery). Contractions in labor were augmented with a standard protocol of oxytocin if progress was slow. Epidural analgesia was used by 51.9%, and the mean duration of usage was 238 ± 205 minutes (range 13–1112). The remaining women used nitrous oxide, transcutaneous electric nerve stimulation, or pethidine, or had local infiltration or a pudendal block for instrument-assisted delivery. All episiotomies were right mediolateral. Minor and uncomplicated perineal lesions were sutured by midwifery staff and house officers. More complicated perineal trauma including third-degree tears were sutured by senior staff, registrars, and consultants. Six women sustained third-degree tears, four of whom were delivered by forceps and two who had episiotomies only. All women were catheterized before assisted vaginal delivery; other indications for catheterization included palpable bladder with inability to void.

There was a high prevalence of irritative urinary symptoms (frequency, nocturia, and urgency) and stress incontinence in pregnancy that decreased after delivery (Table 2). Among 20 women who had stress or urge incontinence before pregnancy, symptoms persisted in 16 during pregnancy and in seven at 12 weeks postpartum. None of the women reported urge and stress incontinence before pregnancy. One woman who had mild symptoms of mixed incontinence before pregnancy had no incontinence in pregnancy or postpartum. All women who had mild incontinence did not need to wear sanitary protection.

Table 2

Table 2

Pregnancy and delivery resulted in significant increases in symptoms of urgency, urge incontinence, and stress incontinence, but there was a decrease in urinary frequency (Table 2). The prevalence of nocturia was unchanged from prepregnancy values. During pregnancy, 240 women (43.7%) reported symptoms of stress or urge incontinence; 224 (93.3%) of these women had onset of symptoms in pregnancy, and 43 (19.2%) of these had symptoms that persisted after delivery. The frequency and severity of incontinence were higher in pregnancy, with 51 women (21.3%) having episodes of incontinence at least once per day. All of these women required sanitary protection and seven (13.7%) felt that it restricted their activities. Postnatally, of 80 women (14.6%) with urinary incontinence, seven (8.8%) had symptoms that were present before and during pregnancy and 30 (37.5%) had onset of symptoms after delivery. The prevalence of de novo urinary incontinence was 5.5%. Postpartum urinary incontinence was not related to obstetric variables.

Pregnancy and delivery were associated with significant increases in defecatory symptoms, mainly postpartum fecal urgency and incontinence of flatus (P < .05; Table 3). Before pregnancy, fecal urgency and incontinence were confined to women with histories of irritable bowel syndrome. In pregnancy, 70 women (12.8%) reported symptoms of fecal urgency or incontinence; 63 (90%) of these had onset of symptoms in pregnancy, and 12 (17.1%) had symptoms that persisted after delivery.

Table 3

Table 3

Postnatally, of 47 women (8.6%) with fecal urgency or incontinence, three (6.4%) had symptoms that were present before and during pregnancy, and 33 (70.2%) had onset of symptoms after delivery. The prevalence of de novo fecal urgency or incontinence was 6%. Women who had cesareans were significantly less likely to report fecal urgency than those who had vaginal deliveries (Table 4), but there was no association of obstetric variables with postpartum incontinence of flatus. Combined urinary and fecal incontinence occurred in 3.6% of postpartum women. Postnatal fecal and urinary symptoms were not related to smoking, race, first-trimester body mass index, striae, hernia, varicose veins, hemorrhoids, or family history of incontinence or prolapse.

Table 4

Table 4

Joint-mobility scores were normally distributed, with higher mobility associated with incontinence of flatus. Symptomatic women had a mean joint-mobility score of 4.7 ± 2.1 (n = 27) and asymptomatic women scored 3.8 ± 2.0 (n = 513) (P = .021). Joint-mobility scores were not related to fecal urgency, fecal incontinence, or urinary incontinence.

The results of the logistic regression analysis determining the relation between obstetric variables and postpartum symptoms are shown in Table 5. Univariate analysis for postpartum symptoms and physical markers of collagen weakness showed no significant differences between groups (except joint-mobility scores), and multiple logistic regression analysis was not done.

Table 5

Table 5

Back to Top | Article Outline


Our study confirmed previous reports of a high prevalence of urinary and fecal incontinence related to pregnancy and childbirth.12,13 None of these women complained voluntarily or sought medical help, which is not surprising because these symptoms are embarrassing and known to be underreported.14 Although data regarding prepregnancy symptoms were retrospective, the low incidence of incontinence agrees with previous data linking parity to incontinence.12,15,16 We chose not to do objective tests because it would have affected the participation and continuation rates and consequently bias the final analysis. There is a poor correlation between urodynamic studies and subjective symptoms in pregnant and nonpregnant women.17,18

Compared with postpartum, the prevalence of all urinary and fecal symptoms was higher during pregnancy. The higher prevalence of stress incontinence in pregnancy might be explained by the mechanical effect of the gravid uterus on the bladder or possibly the effect of hormonal influences on urethral physiology. Thirty women (5.5%) reported de novo postpartum urinary incontinence, which differs from the results reported by Stanton et al19 and Francis,20 who found that stress incontinence rarely appeared for the first time postpartum. However, other authors21,22 have reported that up to 14.1% of women develop de novo urinary incontinence. The prevalence of postpartum stress incontinence cannot be fully explained by trauma secondary to vaginal delivery because 8.4% reported stress incontinence after cesarean. Iosif and Ingemarsson23 reported stress urinary incontinence in 8.8% of women after elective cesarean, suggesting a mechanism other than trauma sustained during vaginal delivery. It was suggested that urinary incontinence after elective cesarean is probably due to detrusor instability resulting from vesical denervation.12

The present study differs from that of Sultan et al,10 who found that no nulliparous women who had anorectal physiology tests and ultrasound were symptomatic antenatally. Six percent of women developed de novo fecal incontinence postpartum, similar to reports by Sultan et al10 (4%) and MacArthur et al13 (4%); however, the latter did not ask about incontinence of flatus. Studies attempting to correlate obstetric variables with postnatal symptoms have not reported consistent findings, which might be due to difficulty in distinguishing the effects of various obstetric factors and their interactions with each other when doing multivariate analyses.12,13,22

Fecal urgency was associated with mode of delivery. Women who had cesareans were significantly less likely to have this symptom compared with those who had spontaneous or instrumental vaginal deliveries. We were unable to show the difference between forceps and vacuum extraction that was shown by Sultan et al,24 who reported that forceps delivery was independently associated with sphincter trauma. MacArthur et al13 identified vaginal deliveries, forceps, and vacuum extractions as independent risk factors for fecal incontinence. These studies implicate pelvic-floor trauma due to vaginal delivery in the etiologies of fecal urgency and incontinence.

The relation between physical markers of collagen weakness and urinary or fecal incontinence has not been investigated previously. Our study found that higher mobility scores were related to incontinence of flatus only. Other physical markers studied were not predictive of postpartum urinary or fecal incontinence. The contribution of family history as a risk factor for developing urinary incontinence was suggested previously,25 but our data did not support this. Case-control evaluations of familial aggregation can be biased by retrospective data collection.26 One's own history of incontinence might be more likely to stimulate information regarding family histories of incontinence.

Although collagen weakness was previously implicated in the pathogenesis of incontinence, the physical markers used in this study could not predict postpartum urinary or fecal incontinence. Either these markers are not representative of collagen weakness, or a larger study with longer follow-up is needed.

Back to Top | Article Outline


1. Burgeson R. Genetic heterogeneity of collagens. J Invest Dermatol 1982;79:25–30.
2. Norton P, Boyd C, Deak S. Collagen synthesis in women with genital prolapse or stress urinary incontinence. Neurourol Urodyn 1992;11:300–1.
3. Ulmsten U, Ekman G, Giertz G, Malmstrom A. Different biochemical composition of connective tissue in continent and stress incontinent women. Acta Obstet Gynecol Scand 1987;66:455–7.
4. Norton PA. Pelvic floor disorders: The role of fascia and ligaments. Clin Obstet Gynecol 1993;36:926–38.
5. Falconer C, Ekman G, Malmstrom A, Ulmsten U. Decreased collagen synthesis in stress-incontinent women. Obstet Gynecol 1994;84:583–6.
6. Keane DP, Sims TJ, Bailey AJ, Abrams P. Analysis of the pelvic floor electromyography and collagen status in premenopausal nulliparous females with genuine stress incontinence. Neurourol Urodyn 1992;11:308–9.
7. Parks AG, Swash M, Urich H. Sphincter denervation in anorectal incontinence and rectal prolapse. Gut 1977;18:656–65.
8. Beersiek F, Parks AG, Swash AM. Pathogenesis of anorectal incontinence: A histometric study of anal sphincter musculature. J Neurol Sci 1979;42:111–27.
9. Calguneri M, Bird HA, Wright V. Changes in joint laxity occurring during pregnancy. Ann Rheum Dis 1982;41:126–8.
10. Sultan AH, Kamm MA, Hudson CN, Thomas JM, Bartram CI. Anal sphincter disruption during vaginal delivery. N Engl J Med 1993;329:1905–11.
11. Beighton P, Solomon L, Soskolne CL. Articular mobility in an African population. Ann Rheum Dis 1973;32:413–5.
12. Wilson PD, Herbison RM, Herbison GP. Obstetric practice and the prevalence of urinary incontinence three months after delivery. Br J Obstet Gynaecol 1996;103:154–61.
13. MacArthur C, Bick DE, Keighley MRB. Faecal incontinence after childbirth. Br J Obstet Gynaecol 1997;104:46–50.
14. Leigh RJ, Turnberg LA. Faecal incontinence: The unvoiced symptom. Lancet 1982;i:1349–51.
15. Thomas TM, Plymat KR, Blannin J, Meade TW. Prevalence of urinary incontinence. BMJ 1980;281:1243–5.
16. Foldspang A, Lam GW, Elving L. Parity as a correlate of adult female urinary incontinence prevalence. J Epidemiol Community Health 1992;46:595–600.
17. Cutner A, Cardozo LD, Benness CJ. Assessment of urinary symptoms in the second half of pregnancy. Int Urogynecol J 1992;3:30–2.
18. Jarvis GJ, Hall S, Stamp S, Millar DR, Jonhson A. An assessment of urodynamic examination in incontinent women. Br J Obstet Gynaecol 1980;87:893–6.
19. Stanton SL, Kerr-Wilson R, Harris GV. The incidence of urological symptoms in normal pregnancy. Br J Obstet Gynaecol 1980;87:897–900.
20. Francis WJA. The onset of stress incontinence. J Obstet Gynaecol Br Emp 1960;67:899–903.
21. Beck RP, Hsu N. Pregnancy, childbirth and the menopause related to the development of stress incontinence. Am J Obstet Gynecol 1965;91:820–3.
22. Viktrup L, Lose G, Rolff M, Barfoed K. The symptom of stress incontinence caused by pregnancy or delivery in primiparas. Obstet Gynecol 1992;79:945–9.
23. Iosif CS, Ingemarsson I. Prevalence of stress incontinence among women delivered by elective caesarean section. Int J Gynaecol Obstet 1982;20:87–9.
24. Sultan AH, Johanson RB, Carter JE. Occult and sphincter trauma following randomized forceps and vacuum delivery. Int J Gynaecol Obstet 1998;61:113–9.
25. Skoner MM, Thompson WD, Caron VA. Factors associated with risk of stress urinary incontinence in women. Nurs Res 1994;43:301–6.
26. Sackett DL. Bias in analytic research. J Chronic Dis 1979;32:51–63.
© 1999 The American College of Obstetricians and Gynecologists