Urinary incontinence (UI) is defined as any involuntary loss of urine. Its prevalence varies from 17% to 55%, with a noticeable relationship between the increase in UI prevalence with increasing age. Currently, Brazil is experiencing a rapid population aging. It is estimated that there are around 21 million Brazilians aged 60 or over, with more than half of them women. Considering current demographic data and UI prevalence in Brazil, it is estimated that approximately 2 million women have UI.
This scenario of elderly Brazilian women with UI is especially worrying due to a few reasons. As part of the aging, there is a loss of muscle mass and strength that negatively influences functionality and increases risk of falls. At the same time, having UI can be a marker of a general poor health condition that can also generate an indirect cause of falls.
The relation between having UI and functionality and falls is not clear in the literature. It is hypothesized that there is a difference in the functional performance of the lower limb, overall muscle strength and risk of falls in elderly women depending on their continence status. A recent systematic review found a higher rate of mortality (hazard ratio = 2.392) in women with UI  when compared to continent ones. Then, the aim of this review is to verify if there is a relationship between each of the following functional variables (strength, balance, postural control, flexibility, gait, general physical performance and falls) with urinary incontinence in elderly women.

This is a systematic review of cross-sectional studies. Electronic search was conducted on  MEDLINE/Pubmed, Embase, CINAHL and Web of Science in September 2022. Two independent reviewers performed the selection process based on titles and abstracts followed by full-text screening. Any disagreement between them was resolved  thru discussion with a third reviewer. Rayyan was used as review management.  
Inclusion/exclusion criteria was based on PICOD scheme: Participants (community-living women aged 60 years or older, with UI and without history of neurological disorders or cognitive impairment); Indicator (correlation, association or regression); Comparison (community-living women aged 60 years or older and without UI and without history of neurological disorders or cognitive impairment); Outcome functional variables (strength, balance, postural control, flexibility, gait, general physical performance and falls); Design (primary full-paper of cross-sectional studies). 
Due to differences in outcome measures between the included studies, a descriptive analysis was performed. Meta-analysis was performed when possible. RevMan version 5.4 software was used. Heterogeneity within each pairwise comparison was evaluated with the I2 value. If the value of I2 is less than 50%, a fixed effects model was used; otherwise, a random effects model was employed. The mean difference (MD) and odds ratio (OR) were considered for continuous and dichotomous variables, respectively.

The electronic search resulted in 653 references, of which 38 went through full-text screening. Six studies were included in the review with a  total of 2549 participants of which 1028 women reported UI. Included studies used different outcome measures for lower limb function (strength, balance, postural control, flexibility, gain, general physical performance). Only 3 studies reported falls history.
In one study, it was found that there was no significant difference in knee extensor and flexor muscle strength (measured in 30-degree and 60-degree position), between women with and without UI. Two studies used the sit-to-stand test, however in different manners.  The 30-second sit-to-stand test found no statistical significant difference in the number of movements performed in 30 seconds. The five times sit-to-stand test also found no difference in the time to perform the test between groups. 
 Postural control was assessed in two articles using a force platform. The analysis was with no statistical significant difference between continence status with displacement of center of pressure (COP) for one of those studies. The second manuscript found a difference between continent and incontinent women in the average speed of COP (p=0.0048) and oscillation area of COP (p=0.0139) with eyes closed. In the dual task test, there was a difference in the oscillation area of COP between women with and without UI, and women with UI had worse results (p=0.0331). However, a joint analysis of the data was not performed because the latter manuscript did not present exact numbers.
Two studies evaluated the gait variable. The first of them used the 6-meter walking  and Time Up and Go (TUG) tests. The results showed the worst value for women with urgency urinary incontinence (UUI) or mixed urinary incontinence (MUI) compared with continent women (p<0.001), but there was no difference between women with stress urinary incontinence (SUI) and continent women. The second manuscript used the 10-meter walking and 6-minute walking tests. In this study, women with UI demonstrated a slower gait speed than continent women in the 10-meter test (p=0.05) but no significant differences between the two groups were found for the 6-minute walking test. 
Balance was assessed by three manuscripts, and it was possible to perform meta-analysis with two of them using the “Four Square Step Test” data (Figure 1).  Fall history was assessed by three studies and all of them were analyzed on a meta-analysis and presented in Figure 2.

The present study identified that  there is a low number of studies aiming to understand physical functioning differences between continent and incontinent women. The studies included in this review used different outcome measures and those studies using the same test, presented the data in different effect measure/variability. This prevented us from pooling data to better understand the whole scenario for the relationship between the studied variables.  
In general, no affirmation can be done regarding physical functioning differences between continent and incontinent women. The meta-analysis runned for balance and fall history showed no effect between groups. However it is important to highlight a small number of studies with a few participants each and the presence of a moderate to high heterogeneity.

There seems to be a difference on gait and displacement of center pressure between continent and incontinent women, however this difference doesn't appear in strength and balance tests as well as fall history. In spite of this, no affirmation can be done once there is a large heterogeneity in the analyzed outcomes to allow robust statements based on qualitative analysis.