First pregnancy and delivery have been associated with a particular risk of stress urinary incontinence (SUI),1–3 although the symptom may temporarily resolve in the puerperium. The presence of SUI 3 months after delivery in primiparous women with symptoms during the preceding pregnancy or postpartum period has been identified as a predictor of symptoms 1 and 5 years later,3,4 but the underlying etiology is not completely understood. There is retrospective epidemiological evidence to suggest that both pregnancy and vaginal delivery imply increased risk of urinary incontinence,1,5 but limited cohort studies have examined the long-term risk of SUI according to the presence or absence of incontinence during pregnancy and/or the puerperal period.3,6
The aim of this longitudinal cohort study was to estimate the impact of onset of SUI during first pregnancy or the puerperal period for the risk of SUI symptoms 12 years later. We also investigated associations between obstetric and other risk factors connected with the first pregnancy and delivery and the risk of SUI symptoms 12 years after the delivery.
MATERIALS AND METHODS
In 1989, a total of 305 primiparous women were included consecutively when they attended the department for their first delivery. A standard questionnaire formulated according to the definitions of the International Continence Society at that time was used when the women were interviewed a few days after the first delivery and again by telephone 3 months later.3 The women were asked 1) whether they had urinary incontinence provoked by physical exertion, 2) whether they had daily incontinence, and 3) whether the incontinence was a hygienic or a social problem.7 Each of these three categories of SUI are considered outcomes in this article. Women with concomitant urge symptoms were not excluded.
Maternal, obstetric, and neonatal data for the first delivery were obtained from the birth registration forms. Five years later an identical questionnaire was mailed to the same 305 women; 278 women responded. Additional questions pertained to more detailed frequency and duration of incontinence symptoms, subsequent pregnancy, delivery, and received incontinence treatment. The validity of the questionnaire and the results have previously been reported.4 Now 12 years after first delivery, a questionnaire was mailed to the 278 women who answered the 3-month and the 5-year questionnaires. Nonresponders 12 years after first delivery were addressed one more time and, finally, one time by telephone. A total of 242 women responded (79%, 242 of 305). One woman didn’t answer the specific question about SUI 12 years after first delivery, so the study sample comprised 241 women. This questionnaire also included questions concerning weight before first pregnancy, height, months of breastfeeding after each delivery, treatment received, level of current pelvic floor muscle training, and treatment desired if symptoms were present. Records concerning subsequent delivery or surgery that could influence the lower urinary tract during the 12-year period were acquired from 23 different hospitals.
Chi-square analyses were performed to evaluate differences between proportions. Independent samples t tests were performed to evaluate differences between means.
From the sample of 241 primiparous women, three cohort subgroups were defined. Group 0 included women without initial stress incontinence before or during the first pregnancy or the first 3-month puerperium (n=146). The 12-year incidence of SUI was based on these 146 women. Group 1 included women with onset of SUI during the first pregnancy (n=66). Group 2 comprised women with onset of SUI during the first 3-month puerperium (n=18). These groups represent the exposures in this article, with group 0 as the reference group. Onset of SUI was unclear in two women, whereas nine women had onset before the first pregnancy, and these 11 women are left out of analyses of these exposures. Variables listed in Table 1 were considered as potential confounders or effect modifiers when comparing group 1 compared with 0 and group 2 compared with 0. The distribution within these variables and the proportion of women with SUI in each category are shown in Table 1.
We preferred to investigate body mass index (BMI) before the first delivery rather than current BMI, because the former was more closely related to the exposure in time. However, these two parameters were highly correlated (Spearman ρ=0.74, P<.01). Effect modification was tested by the Breslow-Day test for homogeneity between odds ratios (OR) after stratified analyses.8 Confounding was evaluated by means of Mantel-Haenszel common ORs and logistic regression analyses.9 Odds ratios were reported with 95% confidence intervals (CIs). All tests were two-sided, and P<.05 was chosen as level of statistical significance. SPSS 12.0 (SPSS Inc, Chicago, IL) was used for statistical analyses.
Informed consent was obtained from each woman. The Danish National Board of Health, as well as the local ethics committee, approved the study.
Mean age during first delivery was 26 years (range 17–41), and mean number of years since the last delivery was 8.6 (range 0–13). Before the first pregnancy and delivery, the mean BMI was 22.1 (range 15.2–36.1). Twelve years after first pregnancy and delivery, a total of 201 women had one to five additional pregnancies, and 187 women one to four additional deliveries (Table 1). The mean weight among the firstborns was 3,352 g (range 1,480–5,400). Mean gestational age was 39.4 weeks (range 33–44). A total of 228 women disclosed that they breastfed for a mean duration of 6.5 months (range 1–30). Additional details about first delivery have been reported elsewhere.4 No difference in SUI rate 3 months after delivery was observed between the groups of responders and nonresponders 12 years after first delivery, whereas responding women were slightly older (26.1 compared with 24.7 years at first delivery, P<.05).
In seven women, surgical procedures that theoretically could affect the continence function during the 12-year period included hysterectomy because of fibroma uteri, bladder extension because of interstitial cystitis, vaginal surgery because of pelvic organ prolapse, corrections of previous episiotomy, anal sphincter correction, and lumbar or sacral herniated disc. Three of these women had SUI 12 years after first delivery.
Twelve years after the first delivery, the prevalence of SUI was 42% (102 of 241). Daily SUI was found in 5.4% (13 of 241), whereas 8.7% (21 of 241) complained of hygienic or social discomfort because of SUI (Fig. 1). The 12-year SUI incidence was 30% (44 of 146). The prevalence of any SUI 12 years after the first pregnancy and delivery was significantly higher in women in group 1 and group 2 compared with women in group 0 (Table 2). The prevalence of daily SUI was higher, although not significantly, in group 1 compared with group 0. The prevalence was significantly higher in group 2. This latter pattern was also found for women with social or hygienic inconvenience because of SUI (Table 2).
In 70 women who had onset of SUI during the first pregnancy or the first 3-month puerperal period but remission 3 months postpartum, a total of 40 (57%) had SUI 12 years later. In 11 women with onset of SUI during the first pregnancy or 3-month puerperal period but no remission 3 months postpartum, a total of 10 (91%) had SUI 12 years later. In 10 women with onset of SUI during the first pregnancy or 3-month puerperal period but no remission 3 months or 5 years postpartum, a total of nine (90%) had SUI 12 years later.
Unadjusted odds ratio for any SUI in group 1 compared with group 0 was 3.0 (95% CI 1.6–5.4) (Table 2). Body mass index before first pregnancy was the only variable actually confounding this relationship. When adjusting for this factor, the odds ratio changed to 3.5 (95% CI 1.8–6.8). No significant effect modification was found. The number of women with daily SUI or women with social or hygienic inconvenience because of SUI was too small to perform adjusted analyses. Among women who had a vaginal delivery, the prevalence of SUI was 33.0% (38 of 115) in group 0 and 60.8% (31 of 51) in group 1, giving an unadjusted odds ratio of 3.1 (95% CI 1.6–6.2). Among women who had a cesarean delivery, the corresponding figures were 19.4% (6 of 31) and 40.0% (6 of 15), giving an odds ratio of 2.8 (95% CI 0.7–10.9), and there was no significant effect modification of delivery mode.
Unadjusted odds ratio for SUI in group 2 compared with group 0 was 8.1 (95% CI 2.5–26.0) (Table 2). Multivariate analyses of this relationship were not performed due to small numbers. All women with SUI onset after first delivery (group 2) had a vaginal delivery.
When comparing the group of women with SUI 12 years after first delivery with those without SUI among all 241 women, higher BMI before first pregnancy and delivery was significantly associated with an increased risk of SUI 12 years after first delivery, whereas breastfeeding more than 6 months after first delivery or a cesarean performed at first delivery significantly decreased this long-term risk (Table 1). Second stage of labor lasting more than 60 minutes was associated with increased risk of long-term SUI but reached only borderline significance (Table 1).
None of the women had surgery because of SUI, and none volunteered information about lifestyle intervention or medication because of SUI, although two women reported pelvic floor muscle training because of incontinence. When asked specifically about pelvic floor muscle training, 69% (160 of 231) confirmed that they performed training at frequencies ranging from less than twice monthly to more than twice daily (10 women did not answer the question). The prevalence of SUI was 45% (72 of 160) and 39% (28 of 71) in the groups of women who did and those who did not report pelvic floor muscle training activity, respectively. This difference was not statistically significant. Eighteen percent (17 of 93) of women with SUI 12 years after first delivery desired treatment if accessible.
During the 12 years of observation after first pregnancy and delivery, the prevalence of women with any SUI increased, whereas the increase in women with symptoms of daily SUI or SUI causing hygienic or social inconvenience was less pronounced (Fig. 1). Onset of any of the three categories of SUI during first pregnancy or puerperium was associated with similar symptoms 12 years later, although not all associations reached statistical significance. First delivery represented a higher risk of inducing long-lasting SUI than first pregnancy, but the impact of different obstetrics risk factors during first delivery attenuated over time in this sample of women.4 A cesarean seemed to be a protective factor with a long-lasting effect. Additional deliveries or surgical procedures during the observation period did not have any influence on any SUI 12 years after first delivery in this population.
The presence of any SUI 3 months after first delivery predicted a very high risk of any SUI after 5 and 12 years (91%), but these data are based on only 11 women. In women with remission of SUI 3 months after first delivery, 57% had symptoms 12 years later. In women without symptoms during first pregnancy or puerperium, 30% developed any SUI during the 12 years. This confirms our previous findings that onset of SUI during first pregnancy or puerperal period carry a dose-response–like risk of inducing instant or delayed SUI that may become chronic and be present 5 and 12 years later.4
The real impact of a cesarean delivery cannot be determined from this study because we did not have a control group of nulliparae, and we did not separate women with an emergency or a planned cesarean.10,11 Although women with a cesarean may have a better protection of the lower urinary tract and pelvic floor compared with those who deliver vaginally,5,12,13 our data showed that, even if women with SUI during first pregnancy had a cesarean delivery, they had a higher prevalence of symptoms (40%) 12 years later than women without SUI during pregnancy who delivered vaginally (33%). However, these results are based on small numbers of women and have to be confirmed in larger longitudinal studies.
Daily SUI reached pregnancy level again 12 years after first delivery but seemed to be more accepted during pregnancy than 12 years after (Fig. 1). Stress urinary incontinence causing social or hygienic inconvenience is now an entity that has been discouraged in epidemiologic studies14 and is no longer included as such in the International Continence Society definitions. Although the type and quality of epidemiologic questionnaires have improved over the years, we decided to maintain this definition for this prospective study for the purpose of comparison over the 12 years of observation.
Although the impact of most obstetric risk factors was not significant, a cesarean during first delivery seemed to protect against long-term SUI. Vacuum extraction, episiotomy, birth weight more than 4,000 g, and second stage of delivery lasting more than 60 minutes significantly increased the risk of any SUI shortly after first delivery,3 and vacuum extraction and episiotomy significantly increased the risk of any SUI 5 years after first delivery.4 All of these variables were associated with SUI in bivariate analyses in the present study, but no association was statistically significant (Table 1). The present study does not have enough power to decide whether these associations are spurious or if they are subject to type-2 error.
The underlying etiology for pregnancy and delivery inducing SUI is not completely understood, especially for long-term symptoms. During pregnancy, hormonal changes or autonomic denervation may play a role, whereas injuries to the muscle, nerve, or connective tissues supporting the lower urinary tract and pelvic floor may be associated with vaginal delivery.4,15–17 Few prospective long-term studies involving objective findings have been conducted. Abnormalities in levator ani muscle magnetic resonance imaging 9–12 months after first delivery have been reported to be more common in women with urinary incontinence than in continent primiparous women and even more common than in nulliparous women.18 Pelvic nerve injury after delivery has been associated with instant or delayed SUI in a 5-year neurophysiological study,19 but reliable long-term neurophysiological data 7 and 15 years after delivery have proven difficult to obtain.20
When evaluating the risk of any SUI 12 years after first delivery in group 1 compared with group 0, overweight before first pregnancy was the only confounder for the risk of any SUI. All eight women with a BMI of 30 kg/m2 or greater before first pregnancy had SUI 12 years after first delivery (Table 1). Some recall bias should be expected because these data were based on weight information obtained with the questionnaire 12 years after first delivery. However, obesity is an established risk factor for the development of urinary incontinence, especially SUI. The impact of prepregnancy obesity on long-term SUI risk has not been reported before, but Burgio et al6 found that, in 523 multiparous women, obesity was related to urinary incontinence in the first year after delivery. Obesity is not only a risk factor for immediate or long-term SUI but also increases the risk of poor surgical outcome when treating women with SUI. When 192 women answered a postal questionnaire 10–18 years after a Burch colposuspension because of genuine SUI, presurgical obesity (ie, BMI of 30 kg/m2 or more) increased significantly the risk of long-term urinary incontinence.21
Breastfeeding for more than 6 months after first delivery reduced the risk of long-term SUI 12 years later in bivariate analyses (Table 1). Burgio et al6 found in 523 multiparous women that breastfeeding significantly increased the risk of urinary incontinence in the first year after the delivery and suggested hormonal changes could explain a transient effect. These different findings may be random, or breastfeeding may have varying impact at different times in a life span. Additionally, we obtained the information about breastfeeding 12 years after the first delivery and some recall bias may be considered.
Stress urinary incontinence remains today an undiagnosed and untreated condition; 12 years after first delivery very few women had received accessible treatment, although 18% of the women with incontinence desired treatment if possible. The estimated number of annual surgical procedures because of SUI during the observation period in Denmark was less than 1,000, and in line with this low estimate none had received continence surgery in this small sample of community-dwelling women with SUI. When questioned directly about pelvic floor muscle training, 69% of the women confirmed they did some form of pelvic floor muscle training at a frequency ranging from less than twice monthly to more than twice daily. To what extent any formalized instruction had taken place was unknown. Stress urinary incontinence was equally present in the group of women who reported pelvic floor muscle training activity as in those without. When Burgio et al6 followed 523 multiparous women 1 year after delivery, they found that pelvic floor muscle training during pregnancy or the postpartum period did not reduce urinary incontinence. These observational studies are, however, not the appropriate way to determine treatment effect, and it is well known that pelvic floor muscle training that is not initially supervised and thoroughly performed regularly has limited protection against SUI. Randomized trials have reported that intensive pelvic floor muscle training during pregnancy may prevent urinary incontinence during the pregnancy and puerperium,22 but the impact of well-performed pelvic floor muscle training during pregnancy on long-term SUI is unknown. When Bo et al23 reinterviewed 47 women with genuine SUI 15 years after they received either home or intensive pelvic floor muscle training, they found disappointing results; long-term adherence with initial training was low and urinary symptoms and severity were not different between exercise groups. Additionally, half of the women had continence surgery with modest effect. Our study provides a naturalistic picture of community-dwelling women with urinary incontinence and highlights the need for better follow-up and management in many women with urinary incontinence.
The strength of this study is the longitudinal design and the long follow-up period. The response rate is very high, and selection bias is minimized. However, the size of the study limits the possibility of performing multivariate analyses to explore differences between subgroups. This was particularly a problem when investigating onset of SUI after the delivery. Also, caution should be exercised when interpreting any insignificant results because of the risk of type-2 error.
We conclude that the onset of SUI during first pregnancy, and especially after first delivery, is significantly associated with symptoms, not only 5 years, but also 12 years later. A cesarean delivery may protect against long-lasting SUI in premenopausal women. Obesity before first pregnancy and delivery seemed to increase the risk of SUI that may become chronic and remain 12 years later. The impact of breastfeeding on the development of SUI is unclear and warrants further investigations. Many women are untreated and could benefit from pre- and postnatal consultation to reduce instant or delayed SUI that remains for years thereafter.
1. Kuh D, Cardozo L, Hardy R. Urinary incontinence in middle aged women: childhood enuresis and other lifetime risk factors in a British prospective cohort. J Epidemiol Community Health 1999;53:453–8.
2. Rortveit G, Hannestad YS, Daltveit AK, Hunskaar S. Age- and type-dependent effects of parity on urinary incontinence: the Norwegian EPINCONT study. Obstet Gynecol 2001;98:1004–10.
3. 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.
4. Viktrup L, Lose G. The risk of stress incontinence 5 years after first delivery. Am J Obstet Gynecol 2001;185:82–7.
5. Rortveit G, Daltveit AK, Hannestad YS, Hunskaar S. Urinary incontinence after vaginal delivery or cesarean section. N Engl J Med 2003;348:900–7.
6. Burgio KL, Zyczynski H, Locher JL, Richter HE, Redden DT, Wright KC. Urinary incontinence in the 12-month postpartum period. Obstet Gynecol 2003;102:1291–8.
7. Abrams P, Blaivas JG, Stanton SL, Andersen JT. The standardisation of terminology of lower urinary tract function. The International Continence Society Committee on Standardisation of Terminology. Scand J Urol Nephrol Suppl 1988;114:5–19.
8. Breslow NE, Day NE. Statistical methods in cancer research. Vol 1. The analysis of case-control studies. Lyon, France: International Agency for Research on Cancer; 1980.
9. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:719–48.
10. Groutz A, Rimon E, Peled S, Gold R, Pauzner D, Lessing JB, et al. Cesarean section: does it really prevent the development of postpartum stress urinary incontinence? A prospective study of 363 women one year after their first delivery. Neurourol Urodyn 2004;23:2–6.
11. 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.
12. Farrell SA, Allen VM, Baskett TF. Parturition and urinary incontinence in primiparas. Obstet Gynecol 2001;97:350–6.
13. Eason E, Labrecque M, Marcoux S, Mondor M. Effects of carrying a pregnancy and of method of delivery on urinary incontinence: a prospective cohort study. BMC Pregnancy Childbirth 2004;4:4. Available at: http://biomedcentral.com/1471-2393/4/4
. Retrieved May 11, 2006
14. Foldspang A, Mommsen S. The International Continence Society (ICS) incontinence definition: is the social and hygienic aspect appropriate for etiologic research? J Clin Epidemiol 1997;50:1055–60.
15. Fitzgerald MP, Mollenhauer J, Hale DS, Benson JT, Brubaker L. Urethral collagen morphologic characteristics among women with genuine stress incontinence. Am J Obstet Gynecol 2000;182:1565–74.
16. Hale DS, Benson JT, Brubaker L, Heidkamp MC, Russell B. Histologic analysis of needle biopsy of urethral sphincter from women with normal and stress incontinence with comparison of electromyographic findings. Am J Obstet Gynecol 1999;180:342–8.
17. Dimpfl T, Jaeger C, Mueller-Felber W, Anthuber C, Hirsch A, Brandmaier R, et al. Myogenic changes of the levator ani muscle in premenopausal women: the impact of vaginal delivery and age. Neurourol Urodyn 1998;17:197–205.
18. DeLancey JO, Kearney R, Chou Q, Speights S, Binno S. The appearance of levator ani muscle abnormalities in magnetic resonance images after vaginal delivery. Obstet Gynecol 2003;101:46–53.
19. Snooks SJ, Swash M, Mathers SE, Henry MM. Effect of vaginal delivery on the pelvic floor: a 5-year follow-up. Br J Surg 1990;77:1358–60.
20. 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.
21. Kjolhede P. Long-term efficacy of Burch colposuspension: a 14-year follow-up study. Acta Obstet Gynecol Scand 2005;84:767–72.
22. Morkved S, Bo K, Schei B, Salvesen KA. Pelvic floor muscle training during pregnancy to prevent urinary incontinence: a single-blind randomized controlled trial. Obstet Gynecol 2003;101:313–9.
23. Bo K, Kvarstein B, Nygaard I. Lower urinary tract symptoms and pelvic floor muscle exercise adherence after 15 years. Obstet Gynecol 2005;105:999–1005.