The aim of this study was to evaluate the association between moderate-severe urinary incontinence (UI), stress UI (SUI), urgency UI (UUI) and sedentary behaviour (SB) in community-dwelling older women.

A cross-sectional study with data from cycle 2005-2006 of the National Health and Nutrition Examination Survey (NHANES) is presented. The NHANES consists of a series of complex and multi-stage surveys on a nationally representative, non-institutionalised U.S. population. Women aged 60 years and over with complete data on objective (accelerometer) and self-reported SB and UI were included. Logistic regression models were fitted to the data. The binary dependent variables in the models were the self-report of moderate-severe UI (leakage of more than just drops of urine volume, a few times a month or more), SUI (leakage or control loss of even a small amount of urine with an activity like coughing, lifting or exercise during the past 12 months) and UUI (leakage or control loss of even a small amount of urine with an urge or pressure to urinate and cannot getting to the toilet fast enough during the past 12 months). SB refers to activities that are performed in a sitting or reclining position and are low in energy expenditure (≤ 1.5 metabolic equivalents [METs]). Objective SB was recorded by an accelerometer type physical activity monitor worn on the hip for seven consecutive days during waking hours. Objective SB variables of interest were: average duration of bouts of SB during the waking day and % time in SB. Self-reported SB variables of interest were: hours per day sitting and watching TV/videos, and sitting while doing usual activities (e.g. work-related, household chores). Both variables were recoded as 6-category and 3-category ordinal variables, according to the distribution.  Other independent variables included in the analysis were: age (categorised as an ordinal variable by 5-year categories), race, body mass index (BMI) calculated as kg/m2, number of vaginal deliveries, smoking habit (never versus current/former smoker), number of alcoholic drinks per week and number of comorbidities. Univariate analysis was undertaken for dichotomous (using Chi-square test), ordinal (through Chi-square for linear trend test) and continuous (Student-t test) variables. Final models for moderate-severe UI, stress UI (SUI) and urgency UUI (UUI) were carried out through logistic regression. The Hosmer-Lemeshow test was applied to check the fit of the models (p>0.05).

Descriptive analysis
The final sample was 459 older women and the prevalence of moderate-severe UI, SUI and UUI was 23.5%, 50.5% and 41.4%, respectively. Most variables had missing data proportions under 5%, however ‘average duration of SB bouts’ was found to have 104 (22.7%) missing values. According to accelerometer data, women spent 64% (SD: 12.2) of their waking time in SB (e.g. sitting or lying) and were sedentary for periods of an average 16.8 minutes. 

Univariate analysis
There was a significant association between SB and duration of SB bouts for older women with UUI. Accelerometer data showed older women with moderate to severe UI and UUI were sedentary for significantly longer proportions of the waking day (UI MD 3.268, p=0.014; UUI MD 2.295, p=0.047), and for older women with UUI the average duration of each bout of SB was 19% longer (MD 3.2 minutes; p=0.001) than the mean SB bout duration overall. 

Multivariate analysis
In multivariate analysis (Table 1), there was a significant association between UUI and the accelerometer-measured duration of SB bouts (OR 1.05, 95% CI:1.01-1.09, p=0.006): longer periods of SB were associated with having UUI. In addition, BMI was positively and significantly associated with all types of UI. Black non-Hispanic race was negatively and significantly associated with SUI. All the models were adjusted by age as well as number of vaginal deliveries and comorbidities. The results of the Hosmer-Lemeshow test showed that the final models were a good fit to the data (p-values of 0.491, 0.903 and 0.399 for moderate-severe UI, SUI and UUI models, respectively). None of the self-reported SB variables or the variable ‘% of time in SB’ were significantly associated with UI type when tested in the multivariate analysis.

•	Longer bouts of SB are associated with UUI in older women.
•	Self-reported SB variables showed no association with any type of UI.
•	BMI is positively associated with moderate-severe UI, SUI and UUI in older women.
•	Black non-Hispanic race is negatively associated with SUI in older women.
•	Risk factors for UI may differ with different types of UI.
•	Future studies are needed to further analyse the pathway between SB and continence status.

Increased SB was associated with UUI in community-dwelling older women. The study highlights the importance of objective measurement of SB as self-reports of SB are unreliable. The results suggest that decreasing time in prolonged sitting may be a target for intervention to reduce UUI in older women.