Urinary incontinence is a common disorder in women, with most epidemiologic studies indicating a prevalence of 25–55% in western societies.1 The wide range in the reported prevalence may be attributed to the likewise wide variety of methodology used to investigate the distribution of disease, as well as large discrepancies in age, study population characteristics, and definitions of incontinence.
Supportive tissue trauma and denervation of the pelvic floor are obstetric events frequently attributed to the development of urinary incontinence.2 An association between childbirth and urinary incontinence has been accepted for several decades, mostly based on data from retrospective and cross-sectional studies. Prospective studies extending beyond the first years after index delivery are scarce,3 generally encompassing a limited period of 1,4,5 to 3,6 years postpartum or with an unsatisfying loss to follow-up.7,8 Thus the causal link between childbirth and pervasive urinary incontinence has not been described in the long term, and a large number of factors associated with an increased risk of lower urinary tract symptoms later in life, such as menopause,9 obesity,10 smoking,10 and pelvic surgery,11 make attempts to determine the association challenging. The aim of this prospective observational cohort study was to assess the long-term effects of childbirth on urinary incontinence after the first delivery.
During a 10-week period starting in April 1995, 440 primiparous women were delivered at the Department of Obstetrics and Gynecology, Danderyd Hospital, Stockholm, Sweden. Subjects scheduled for an elective cesarean delivery (n=65), with inadequate knowledge of the Swedish language (n=18), carrying a duplex pregnancy (n=8), or primarily delivered by acute cesarean (n=45) were excluded from the study. The final analyses were restricted to women with first-time vaginal deliveries (index delivery). During the follow-up period, subjects were excluded from analysis after a subsequent cesarean delivery or urinary incontinence surgery.
Subjects with normal pregnancies and deliveries were primarily delivered by midwives, although the attending physician, (a resident or consultant obstetrician) was involved in all cases of complications or when an instrumental delivery was called for. All subjects entered the study at the postpartum maternity ward and completed a standardized questionnaire on bowel and lower urinary tract symptoms. Some of these patients have previously been described regarding anorectal dysfunction at short-term.12 Follow-up, using an identical questionnaire administered and returned by postal mail, was completed at 5 months, 9 months, 5 years, and 10 years after the first delivery. Obstetric data were primarily retrieved from partograms and hospital charts, but questions on subsequent deliveries and medical history was added to the follow-up questionnaire. Perineal tears were classified according to the International Classification of Disease (ICD) 9th and 10th revisions nomenclature.13
The questionnaire on bladder dysfunction was modeled from the Cleveland Clinic Incontinence Score,14 and included the following questions: “Do you experience sudden desires to void urine that are difficult to hold back?” “Do you experience involuntary loss of urine at physical activities?” and “Do you experience bladder emptying difficulties?” A response of “less than once per week” was considered mild symptoms whereas response alternatives “more than once per week” and “daily” were considered moderate-severe symptoms. A nonresponse to any of the questions in the questionnaire was not considered in the analyses.
Statistical analysis was performed using STATISTICA software (StatSoft Inc., Tulsa, OK). A sample size calculation determined that a sample of at least 200 subjects was necessary to demonstrate a 10–30% increase in prevalence of stress urinary incontinence symptoms (primary outcome variable) with α<.05 and β=.8, based on the assumption that premenopausal women have a 10% prevalence of symptoms. The Wilcoxon signed rank test was used to compare ordinal and continuous numerical data. Multivariable logistic regression analyses were used to assess the effect of vaginal delivery on symptoms of urinary incontinence and determine the interaction of covariates. Relative risk (RR) was estimated with 95% confidence intervals (CIs). A P<.05 was considered significant for all analysis.
The study was approved by the Local Ethics Committee at Karolinska Hospital, Karolinska Institute, Stockholm, Sweden. All patients and control subjects gave their informed consent to participate in the study after receiving written and oral information.
After the first vaginal delivery, 304 of 309 primiparous women entered the study (98%). At the 10-year follow-up, 246 subjects (81%) participated, although 11 subjects were excluded from analysis due to cesarean deliveries before the 5-year follow-up, and three subjects were excluded due to cesarean deliveries before the 10-year follow-up. Of the participating subjects, 232 of 246 subjects reported only having had vaginal deliveries, although an additional three subjects were excluded due to urinary incontinence surgery, leaving 229 patients eligible for the 10-year evaluation.
Mean age (±standard deviation) at first delivery was 29.9 years (±4.1), mean maternal weight at entry of first pregnancy was 63.7 kg (±9.6), and mean fetal weight at index delivery was 3,500 g (±466). Somatic disease requiring medication at the time of the study was limited to two patients with asthma, one with sarcoidosis, one with malaria, one with inflammatory bowel disease, and one patient with a neurosis. Mean age in subjects not responding to the survey was 30.1 years (±6.2).
Subsequent obstetric history 10 years after index delivery showed that median parity was 2 (range 1–4), a history of obstetric perineal rupture grade I-II was reported by 235 patients (76%), and perineal rupture grade III-IV had occurred in 43 patients (13%). Episiotomy had occurred in 14 patients (5%), and a history of instrumental delivery was reported by 25 patients (11%).
At the end of follow-up, conservative treatment of urinary incontinence (including physiotherapy and pelvic floor exercises) was reported by four subjects (2%), surgical treatment of urinary incontinence by three subjects (1%), and 13 subjects (6%) reported daily use of pads due to urinary incontinence 10 years after index delivery. None of the patients had undergone pelvic organ prolapse surgery at 10 years follow-up, and none reported anticholinergic treatment.
The frequency of stress urinary incontinence, as well as urinary urgency, increased significantly at 9 months, 5 years, and 10 years follow-up compared with index delivery (Table 1). The effects of vaginal childbirth on symptoms of stress urinary incontinence, urinary urgency, and bladder emptying difficulties are presented in Table 2. The RR for stress urinary incontinence and urinary urgency increased significantly after index delivery regardless of age (Table 2). Bladder emptying difficulties did not increase significantly until 10 years after index delivery (Table 2).
At multivariable regression analysis on obstetric covariates at first or subsequent deliveries, we found no significant association with the development of stress urinary incontinence or urinary urgency at 10 years follow-up (Table 3). At multivariable analysis (adjusted for age and parity), stress urinary incontinence symptoms at 9 months and 5 years follow-up were independently associated with the presence of symptoms at 10 years (RR 13.3, 95% CI 3.9–33.1 and RR 14.1, 95% CI 2.5–18.8, respectively). We found no other predictors of outcome in lower urinary tract symptoms or treatment thereof at multivariable analysis.
When considering symptoms as mild or moderate-severe, there were significant increases in prevalence of stress urinary incontinence and urinary urgency at 10 years follow-up compared with baseline, as well as increasing incidence rates for the 10-year period after index delivery compared with the 10-year preceding the index delivery. The RR (adjusted for maternal age and parity) for moderate to severe urinary incontinence increased significantly 10 years after first delivery (RR 5.8, 95% CI 1.2–33.7). Figures and rates are presented in Table 4.
This long-term prospective observational study describes the natural progression of urinary incontinence and urinary urgency after first delivery and suggests a strong association between vaginal delivery and increasing prevalence of urinary urgency and stress urinary incontinence symptoms 10 years after the first delivery. Although the majority of symptomatic patients in our study experienced mild to moderate complaints, the incidence rate of urinary incontinence episodes increased five to six times after vaginal delivery. When covariates were added to an adjusted multivariate risk model, we found no explanation for this increase other than a direct causality between vaginal delivery and urinary incontinence. These data corroborate results from cross-sectional studies where parity is an important risk factor for symptoms of urinary incontinence in fertile women.15
In agreement with Rortveit et al,15 our multivariate analysis showed that the number of vaginal deliveries was of limited importance for the outcome, but rather that the first vaginal birth per se was associated with incontinence. Viktrup et al3 found that the presence of urinary incontinence 3 months after delivery was predictive of long-lasting symptoms. We were unable to confirm a correlation between repeat vaginal deliveries and the risk of permanent symptoms, even when considering mild symptoms separately. Instead, incontinence symptoms at short-term follow-up (9 months) were strongly predictive of symptoms 10 years after the index delivery, suggesting that short-term continence status may in part predict long-term outcome. These findings substantiate reports suggesting that the first vaginal delivery is associated with the highest risk of loss of urethral support16 and urinary incontinence.3
Anal sphincter rupture, perineal tears or instrumental delivery used as a proxy for severe obstetric trauma to the lower urinary tract did not correlate with the presence of urinary incontinence 10 years after the first delivery. Instrumental delivery has been shown to have injurious urinary tract side-effects in prospective studies;17 however, the detrimental effects of vaginal delivery may not be limited to tears or ruptures of supportive tissues. Instead our data support the considerable evidence indicating that nondisruptive tissue trauma to the pelvic floor may be a major contributor to the complicated, and multifactorial, pathoetiologic mechanisms behind stress urinary incontinence and other pelvic floor disorders.18 This may be explained by stretching of the tissues or ischemic trauma to the distal branches of the pudendal nerve, causing denervation of the intrinsic urethral sphincter.18
Age and degenerative changes of the bladder are often promoted as key risk factors for urinary urgency.19 However, in this presumed premenopausal population, age at the time of first delivery was not associated with an increased risk of urinary incontinence or urgency. The pattern and distribution of urinary urgency was similar to that of stress urinary incontinence, showing pronounced adverse alterations after the first delivery compared with baseline. These data make us conclude, that although the disease mechanisms of urinary urgency differ from those of stress urinary incontinence, vaginal delivery seems to have a major effect on the prevalence of urgency as well. One may speculate that nondisruptive tissue trauma, especially nerve damage, may have a more pronounced effect on de novo urgency symptoms.
The most dramatic changes in bladder function were observed in the range of mild-moderate symptoms. This was reflected by the low rate of surgical interventions over the study period. We could not confirm that maternal age at the time of first delivery was associated with an increased risk for urinary incontinence surgery, as has been suggested.20 Future life events such as menopausal estrogen depletion and periurethral atrophy may, however, cause symptoms to worsen21 and increase demands for surgical correction later in life.
At the time the study was initiated there were no validated Swedish-language bladder function questionnaires in use; we therefore modeled the questionnaire after a validated anal incontinence score. Although not formally validated, we believe that the consistency of our collected data over the 10-year observational period reflects the reliability and comprehensibility of the questionnaire in use. It may be considered a weakness of our study that our primary outcome measure was based on self-reported symptoms, with no objective verification of disease. Stress urinary incontinence is, however, a disorder defined by self-experienced symptoms,1 and the predictive value of urodynamic investigations is limited in stress urinary incontinence.22
In retrospect it may have been preferable to administer the questionnaire before a major life event, such as delivery of the first child, to avoid any possible effects on the response. The conclusions and internal validity of the present study are, however, strengthened by the prospective study design, the duration of follow-up, the high inclusion and response rate, as well as the homogenous study group. To maintain internal validity we excluded women from further analysis after cesarean delivery. Although this limits our ability to generalize our finding to women with both modes of delivery, it increases our ability to make a causal inference.
Urinary incontinence also exists in nulliparous women, and we recognize that other factors, such as a hereditary predisposition,23 menopausal loss of estrogen,9 and pelvic surgery,24 may have a major influence on the development of bladder and urinary dysfunction. Thus, the development of urinary incontinence is likely to be multifactorial, and the lack of a nonparous control group is an obstacle when trying to isolate the effects of ageing. It is, however, implausible that the observed alterations in lower urinary tract function should be attributed to the ageing process, because most of the adverse effects were apparent already 6–9 months after the index delivery. Furthermore, in this cohort of healthy women, most subjects would be premenopausal 10 years after their first delivery, and none had undergone other life events associated with lower urinary tract dysfunction, such as hysterectomy, pelvic organ prolapse surgery, or pelvic organ radiotherapy. Our 10-year longitudinal results substantiate the common belief that vaginal delivery is a key adverse event and a part of the causal pathway for urinary incontinence and urinary urgency in many women. From a women's' health perspective, future studies need to confirm whether cesarean delivery is associated with a lesser need for surgical correction of urinary incontinence and other benign pelvic floor diseases compared with vaginal delivery.25
1. Abrams P, Cardozo L, Khoury S, Wein AE. News from the 3rd International Consultation on Incontinence. Inpharma 2004;1:5.
2. Snooks SJ, Badenoch DF, Tiptaft RC, Swash M. Perineal nerve damage in genuine stress urinary incontinence: an electrophysiological study. Br J Urol 1985;57:422–6.
3. Viktrup L. The risk of lower urinary tract symptoms five years after the first delivery. Neurourol Urodyn 2002;21:2–29.
4. McKinnie V, Swift SE, Wang W, Woodman P, O'Boyle A, Kahn M, et al. The effect of pregnancy and mode of delivery on the prevalence of urinary and fecal incontinence. Am J Obstet Gynecol 2005 193;193:512–7.
5. Schytt E, Lindmark G, Waldenstrom U. Symptoms of stress incontinence 1 year after childbirth: prevalence and predictors in a national Swedish sample. Acta Obstet Gynecol Scand 2004;83:928–36.
6. Bahl R, Strachan B, Murphy DJ. Pelvic floor morbidity at 3 years after instrumental delivery and cesarean delivery in the second stage of labor and the impact of a subsequent delivery. Am J Obstet Gynecol 2005;192:789–94.
7. Dolan LM, Hosker GL, Mallett VT, Allen RE, Smith AR. Stress incontinence and pelvic floor neurophysiology 15 years after the first delivery. BJOG 2003;110:1107–14.
8. Macarthur C, Glazener C, Lancashire R, Herbison P, Wilson D, Grant A. Faecal incontinence and mode of first and subsequent delivery: a six-year longitudinal study. BJOG 2005;112:1075–82.
9. Falconer C, Ekman-Ordeberg G, Ulmsten U, Westergren-Thorsson G, Barchan K, Malmstrom A. Changes in paraurethral connective tissue at menopause are counteracted by estrogen. Maturitas 1996;24:197–204.
10. Hannestad YS, Rortveit G, Daltveit AK, Hunskaar S. Are smoking and other lifestyle factors associated with female urinary incontinence? The Norwegian EPINCONT Study. BJOG 2003;110:247–54.
11. Borstad E, Rud T. The risk of developing urinary stress-incontinence after vaginal repair in continent women: a clinical and urodynamic follow-up study. Acta Obstet Gynecol Scand 1989;68:545–9.
12. Zetterstrom J, Lopez A, Anzen B, Norman M, Holmstrom B, Mellgren A. Anal sphincter tears at vaginal delivery: risk factors and clinical outcome of primary repair. Obstet Gynecol 1999;94:21–8.
13. World Health Organization. International classification of diseases and health-related problems. Tenth revision, ICD10. Geneva (Switzerland): WHO; 1992.
14. Jorge JM, Wexner SD. Etiology and management of fecal incontinence. Dis Colon Rectum 1993;36:77–97.
15. Rortveit G, Daltveit AK, Hannestad YS, Hunskaar S, Norwegian EPINCONT Study. Urinary incontinence after vaginal delivery or cesarean section. N Engl J Med 2003;348:900–7.
16. Dietz HP, Bennett MJ. The effect of childbirth on pelvic organ mobility. Obstet Gynecol 2003;102:223–8.
17. Liebling R, Swingler R, Patel R, Verity L, Soothill P, Murphy D. Pelvic floor morbidity up to one year after difficult instrumental delivery and cesarean section in the second stage of labor: a cohort study. Am J Obstet Gynecol 2004;191:4–10.
18. Snooks SJ, Setchell M, Swash M, Henry MM. Injury to innervation of pelvic floor sphincter musculature in childbirth. Lancet 1984;2:546–50.
19. Harris RL, Cundiff GW, Coates KW, Bump RC. Urinary incontinence and pelvic organ prolapse in nulliparous women. Obstet Gynecol 1998;92:951–4.
20. Carley ME, Turner RJ, Scott DE, Alexander JM. Obstetric history in women with surgically corrected adult urinary incontinence or pelvic organ prolapse. J Am Assoc Gynecol Laparosc 1999;6:85–9.
21. Fantl JA, Cardozo L, McClish DK. Estrogen therapy in the management of urinary incontinence in postmenopausal women: a meta-analysis. First report of the Hormones and Urogenital Therapy Committee. Obstet Gynecol 1994;83:12–8.
22. Sandvik H, Hunskaar S. The epidemiology of pad consumption among community-dwelling incontinent women. J Aging Health 1995;7:417–26.
23. Buchsbaum GM, Duecy EE, Kerr LA, Huang LS, Guzick DS. Urinary incontinence in nulliparous women and their parous sisters. Obstet Gynecol 2005;106:1253–8.
24. Versi E, Cardozo L. The short-term effect of radical hysterectomy on urethral and bladder function. Br J Obstet Gynaecol 1987;94:822–3.
© 2006 The American College of Obstetricians and Gynecologists
25. van Brummen H, Bruinse H, van de Pol G, Heintz A, van der Vaart C. The effect of vaginal and cesarean delivery on lower urinary tract symptoms: what makes the difference? Int Urogynecol J Pelvic Floor Dysfunct 2006 Apr 21;Epub ahead of print.