Hypothesis / aims of study
The relative importance of pregnancy per se, vaginal delivery, and the potential protective effect of cesarean section on the prevalence of urinary incontinence (UI) in the long-term are still unresolved and controversial issues. The aim of this study was to compare the age-dependent prevalence and severity of different aspects of UI in women aged between 40 and 65 years after one pregnancy and one vaginal delivery, and the attributable risk reduction of UI with C-section.
Study design, materials and methods
Three randomly selected national cohorts (nulliparous women, one-para women with a vaginal birth and one-para women with cesarean delivery, n = 14 335) were studied. Nulliparous women were obtained by Statistics Sweden. One-para, vaginally and cesarean delivered women were recruited from the Swedish Medical Birth Register and surveyed 20 years after birth. Self-administered, validated questionnaires for different aspects of UI, using the IUGA/ICS definitions (1) and Sandvik’s severity index (2), were used. The effect of one pregnancy and one vaginal delivery was analyzed with one-to-one matched cohorts. Matching sought to balance for body mass index and age, shown to be strong confounders of UI (3).
Results
The logistic regression model showed that the age-related gap for UI between nulliparous and 1-para women delivered vaginally or by cesarean section was constant between parallel trajectories that spanned between the ages of 40 and 65 years (Figure 1). Pregnancy increased the overall prevalence of UI from 20.1% to 30.1% (OR1.71, 95%CI 1.43-2.05; p<0.0001) and vaginal delivery increased the overall UI prevalence from 30.1% to 43.0% (OR1.75, 95%CI 1.49-2.05; p<0.0001). Vaginal delivery skewed parameters for severity of UI towards more severe forms. Moderate to severe UI after vaginal birth increased from 12.7% to 19.5% (OR1.67, 95%CI 1.35-2.07; p<0.0001), whereas pregnancy did not alter moderate to severe UI (10.6 versus 12.7%, OR1.25, 95% CI 0.98-1.59; p=0.067). The derived protective effect of C-section, i.e. bypassing vaginal delivery, amounted to a 30% reduction in UI (OR 0.57, 95%CI 0.49-0.67; p<0.0001) and 35-43% reduction of more severe forms of UI.
Interpretation of results
The results of this study contradict conclusions from some earlier cross-sectional epidemiological reports that have concluded that the protective effect of cesarean section is temporary, restricted to milder forms of UI, or even non-existent.