Authors' response to: Prevalence of urinary incontinence 20 years after childbirth in a national cohort study in singleton primiparae after vaginal or caesarean delivery


Authors' Reply


We thank Drs Onyeka and Ladanchuk for their interest in our paper, published recently in BJOG.[1] We agree entirely that information regarding different types of urinary incontinence (UI) after childbirth is of importance; however, as stated in our article,[1] our aim was to present the prevalence of UI, based on the International Continence Society definition, 20 years after one delivery, and in particular analyse the possible excess risk of UI after vaginal birth compared with caesarean section, and to evaluate some other important risk factors for UI. As the article stands, it already contains a substantial amount of information, and is quite lengthy, which is why we refrained from presenting data on UI subtypes, severity and impact at that stage. In order to present data on UI subtypes and their influence on quality of life in an adequate manner, we have chosen to present this data separately.

We have already analysed our data from the SWEPOP study with regard to the prevalence, severity and impact associated with the three main subtypes of UI (stress urinary incontinence, urge urinary incontinence and mixed urinary incontinence) 20 years after vaginal delivery or caesarean section. In this section of the SWEPOP study we will present data on severity of UI, significant UI and the subjective perception of UI (evaluated as the degree of bother), and also grouped according to UI subtype. We will also present data on help-seeking behaviour.

We entirely agree with Onyeka and Ladanchuk that it is well established that injury to the pelvic floor occurring as a result of vaginal birth is the single most common causal factor in the etiology of stress urinary incontinence (SUI).[2, 3] We also agree that the etiology of urge urinary incontinence (UUI) relating to childbirth is less well understood. Data on urge urinary incontinence and mixed urinary incontinence (MUI) after childbirth are conflicting.[3] Most cross-sectional and cohort studies with short or 1–year and longer follow-up could not demonstrate a significant difference in the prevalence of UUI and MUI between the two modes of delivery.[3] However the data available has to be interpreted with caution as the number of women who gave birth by caesarean section in the cohort studies cited in this review with 1-year and longer follow-up was low. In the EPINCONT cross-sectional study[2] that Onyeka and Ladanchuk refer to, the difference in prevalence between vaginal birth and caesarean section was significant for SUI, but not for UUI and MUI. This may have resulted from a lack of power, as the prevalence of UUI in the EPINCONT study was comparatively low (2%), and the population was heterogeneous for parity (1–4), women were younger (55% <40 years) and the fraction of women with caesarean sections only was small (6%).[2] Therefore, it is not possible to conclude whether or not UUI is linked or not linked with vaginal delivery from the EPINCONT study, and so we do not agree with Onyeka and Ladanchuks' statement that the etiology of UUI is not usually linked to childbirth. We are of the opinion that more data are necessary before such a statement can be made. We believe that our article to be published concerning data on the prevalence, severity and associated impact of the three main subtypes of UI (SUI, UUI and MUI) 20 years after vaginal delivery or caesarean section will provide further valuable information in this respect.

In the SWEPOP study, caesarean section performed before the onset of labour was denoted as an elective caesarean section, and caesarean section performed during labour was denoted as an acute caesarean section. We maintain that our analysis shows that there is no statistical difference in the prevalence of late UI between these two groups of women. We agree with Onyeka and Ladanchuk, however, that it would have been of great interest to investigate the late prevalence and risk of UI depending on whether the surgery was performed before or during the second stage of labour. Such information is regrettably not registered in The Swedish Medical Birth Registry, and therefore was not available for analysis. This limitation of the Swedish Medical Birth Register was clearly stated in the article.