Longitudinal study of cognitive decline among women with and without urinary incontinence.

backgroundCross-sectional and short-term cohort studies have demonstrated an association between urinary incontinence and dementia, as well as lower performance on cognitive testing. The Health and Retirement Study, a longitudinal study of community-dwelling older adults, offers an opportunity to assess the temporal association between these conditions because it included an assessment of incontinence symptoms and biennial assessments of cognitive function.

objectiveThis study aimed to evaluate if urinary incontinence before the age of 70 years had an effect on changes in cognitive function among women participating in the Health and Retirement Study. STUDY

designThis secondary analysis included data from female respondents in the Health and Retirement Study aged 58 to 67 years with ≥2 cognitive assessments. Urinary incontinence was defined as any involuntary loss of any urine in the preceding 12 months. A control group without incontinence was reweighted for better comparability using coarsened exact matching for age and comorbidities. Validated methods, including neuropsychological test data, estimated a memory score and dementia probability for each participant biennially. Coprimary outcomes were the changes in memory score and dementia probability. Linear regression models were used to estimate the association of urinary incontinence with change in memory score and dementia probability, adjusting for baseline demographics and comorbidities. A subgroup analysis was performed to assess the effects of urinary incontinence frequency on these outcomes. The infrequent subgroup reported <15 days of leakage per month and the frequent subgroup reported ≥15 days of leakage per month.

resultsAmong eligible female respondents, 40.6% reported urinary incontinence between the ages of 58 and 69 years. Baseline memory scores and dementia probability were similar between those with urinary incontinence (n=1706) and controls (n=2507). Memory score declined significantly in both cohorts, indicating poorer memory over time (-0.222 among those with incontinence [95% confidence interval, -0.245 to -0.199] vs -0.207 in controls [95% confidence interval, -0.227 to -0.188]). The decline of memory score was not statistically significantly different between cases and controls (mean difference, -0.015; 95% confidence interval, -0.045 to 0.015). Dementia probability increased significantly in both groups, indicating a greater probability of developing dementia by 0.018 among those with incontinence (95% confidence interval, 0.015-0.020) and by 0.020 among controls (95% confidence interval, 0.017-0.022). The change in dementia probability was not significantly different between groups (mean difference, -0.002; 95% confidence interval, -0.006 to 0.002). Frequent urinary incontinence was reported in 105 of 1706 (6%) of those with urinary incontinence. Memory score declined and dementia probability increased with time (P<.001) in frequent and infrequent urinary incontinence subgroups. There was no dose-response relationship.

conclusionMeasures of cognitive performance declined during approximately 10 years of observation. The changes in performance were not associated with the presence of urinary incontinence in the participants' younger years.