Urinary incontinence in rural older women: prevalence, incidence and remission

Abstract

Objectives: To determine the prevalence, incidence, and remission rates of urinary incontinence in a large group of older women over a 6-year time span and to assess factors associated with incontinence incidence and remission.

Design: Longitudinal cohort study.

Setting: Two rural counties in Iowa.

Participants: 2025 women aged 65 years or older residing in rural Iowa, enrolled in the Iowa 65+ Rural Health Study of EPESE (Establishment of Populations for Epidemiologic Studies of the Elderly) were interviewed in person annually for 6 years; specific responses to queries about urinary incontinence were given at baseline, 3-, and 6-year intervals.

Measurements: Conditional multivariate logistic regression analysis was done to assess the relationship between incontinence symptoms and various factors previously found to be related to incontinence.

Results: The baseline prevalence of urge incontinence was 36.3%, and of stress incontinence it was 40.3%. For urge incontinence, the 3-year incidence and remission rates between the third and sixth years were 28.5% and 22.1%, respectively. For stress incontinence, the 3-year incidence and remission rates between years 3 and 6 were 28.6% and 25.1%, respectively. Seventy-six percent and 84% of women who reported no urge or stress incontinence, respectively, at the baseline interview were continent at both follow-up interviews. The only significant factors related to changes in incontinence status were age, which was associated with an increased incidence of urge incontinence (OR 1.11, P = .017, 95% CI 1.019-1.203), and improvement in activities of daily living, which was associated with a increased remission of urge incontinence (OR 0.50, P = .015, 95% CI 0.28-0.9)

Conclusion: In some older women, urinary incontinence is a dynamic state, with women moving back and forth along a continuum between continence and incontinence. These results are tempered by limitations of the study, which include its questionnaire design and lack of ability to detect potential treatment effect.