Pelvic floor dysfunction depending on mode of delivery : clinical and epidemiological aspects

Abstract

Objective: To study pelvic floor disorders in relation to mode of delivery using clinical and epidemiological methods; to compare the prevalence and risk of lower urinary tract symptoms (LUTS) in healthy primiparous women in relation to vaginal (VD) or elective cesarean (CD) delivery nine months after delivery; to estimate the effect of delivery on urinary and anal incontinence 10 years after first childbirth in relation to mode of delivery and to assess the influence of parity and obstetrical events; to estimate the risks of stress urinary incontinence (SUI) and pelvic organ prolapse (POP) surgery related to delivery mode at long-term follow-up; and to evaluate the influence of age at first childbirth on the risks of surgically managed SUI and POP.

Methods: Paper I is a clinical study of 435 subjects, with prospectively collected data based on self-reported questionnaires and medical charts. Paper II is a cross-sectional comparative study of 395 subjects, and Paper III and IV are nationwide cohort studies including over 90,000 subjects, based on data from the Swedish Medical Birth Register and the Swedish Inpatient Register. Statistical analysis was performed on the clinical cohorts, fulfilling a power criteria of α<0.05, β=0.8, using non-parametric statistics, correlation coefficients and logistic regression. Incidence rates, Hazard Ratios (HR) and NNH were calculated on the population-based cohorts.

Results: VD was associated with a significantly increased prevalence and risk of LUTS nine months after childbirth when compared to elective CD (RRSUI 8.9 95%CI 1.9-42). De Novo SUI was likewise more prevalent after VD. The protective effect of CD was absent in subjects reporting SUI before pregnancy, symptoms before pregnancy or at three months follow-up (RRSUI 3.9 95%CI 1.7-8.5) were independent risk factors for persistent symptoms at nine months follow-up, with a higher prevalence of SUI after VD. At 10 years follow-up were urinary and anal incontinence symptoms significantly more common following spontaneous VD. Women with an obstetrical history of anal sphincter injury carried a substantially increased risk of gas incontinence (OR 3.1 95%CI 1.5-8.9). However, CD is not associated with a major reduction of urinary incontinence symptoms and an association between delivery mode and anal incontinence could not be confirmed. A significantly increased risk of SUI and POP surgery later in life was seen in women with only VD compared to only having had CD (HRSUI 2.9 95%CI 2.4-3.6, HRPOP 9.2 95%CI 7.0-12.1). The increased risk of surgically managed pelvic floor disorders persisted for the three decades of follow-up and was especially pronounced in multiparous women. An age dependent effect was seen with higher age at first delivery associated with higher risk and incidence rate of subsequent pelvic floor surgery in both delivery groups. The protective effect of CD remained in all age categories, predominantly for POP in women ≥ 30 years of age at first childbirth.

Conclusions: The studies in this thesis provides clinical and epidemiological evidence that obstetrical intervention at the time of childbirth may to some extent prevent SUI and POP later in life, particularly in multiparous women and women over 30 years of age at first delivery. However, the protective effects of CD with regard to pelvic floor disorders must be weighed against postpartum maternal and neonatal morbidity associated with the procedure.