The effect of childbirth on urinary incontinence: a matched cohort study in women aged 40–64 years

Abstract

The relative impact of age, pregnancy and vaginal delivery on urinary incontinence is still an unresolved issue that involves the controversial question about the protective effect of cesarean delivery. The purpose of this study was to estimate and compare the effect size of 1 pregnancy, 1 vaginal delivery, and the derived protective effect of cesarean delivery for different aspects of urinary incontinence in women 40-64 years old, all 20 years after birth. This Swedish nationwide matched cohort study involved 14,335 women. Data from 3 restricted, randomly selected, source cohorts of (1) nulliparous women who were unexposed to childbirth (n=9136), (2) primiparous women who had experienced cesarean delivery and who had been exposed to 1 pregnancy (n=1412), and (3) primiparous women who had been exposed to 1 pregnancy followed by vaginal delivery (n=3787) were retrieved from The Swedish Medical Birth Register and Statistics Sweden and surveyed in 2008 and 2014, respectively. Parous women were all assessed 20 years postnatally. One-to-one matching with an interval for pairing of 3 years and 3 body mass index units was used in women 40-64 years old with information about body mass index (kilograms/square meters) and urinary incontinence. The procedure succeeded in 2630 of 2635 women (99.8%) and resulted in an adequate distribution of age and body mass index between groups. The surveys used a postal- and an internet-based questionnaire with validated questions for various aspects of urinary incontinence. Fisher's exact test and the Mann-Whitney U test were used for comparisons between matched groups; trend was analyzed with Mantel-Haenszel statistics. Predicted, age-related values of different aspects of urinary incontinence were obtained by logistic regression analysis. Pregnancy increased the prevalence of urinary incontinence from 20.1-30.1% (odds ratio, 1.71; 95% confidence interval, 1.43-2.05; P<.0001]. Urinary incontinence increased further after vaginal delivery to 43.0% (odds ratio, 1.75; 95% confidence interval, 1.49-2.05; P<.0001); "moderate" and "severe" urinary incontinence increased from 12.7-19.5% (odds ratio, 1.67; 95% confidence interval, 1.35-2.07; P <.0001). There was a parallel increase in urinary incontinence from 40-65 years of age in nulliparous and vaginally and cesarean delivered women. Cesarean delivery, compared with vaginal delivery, was associated with a 30.0% reduction of urinary incontinence (P<.0001) and a 35-52% reduction of more severe grades of urinary incontinence (P<.0001) and was unaffected by age. Both pregnancy and vaginal delivery incurred an increased risk of urinary incontinence in the long term. The age-related gap for urinary incontinence between nulliparous and primiparous women who were delivered by vaginal delivery or cesarean delivery was constant between parallel trajectories that spanned ages from 40-64 years. The calculated protective effect of cesarean delivery was unaltered and significant during the same age interval.