Physical exercises, professional or not, generate impacts that interfere with the pelvic floor musculature. It is still controversial as to whether this impact strengthens or weakens the tone or changes muscle function for determining the protection, onset, or worsening of urinary incontinence (UI). Since it has become a subject of great interest in recent years due to the greater commitment of women to sports and increased data on the pelvic floor in this population, and after a publication of a systematic review on the prevalence of UI in women athletes [1], there are now several publications on this subject matter. Through an updated systematic review, our objective is to determine the prevalence of UI in women who exercise, according to the impact of various sports on the pelvic floor.

A systematic review with meta-analysis was conducted searching PubMed, Cochrane Library, and Lilacs databases until July 31, 2020. The search strategy was utilized with the keywords "Pelvic floor disorders" or "Urinary incontinence" and "Athletes" or "Sport" similarly to the previous publication, with no language restriction. The eligibility criteria included studies with women who practiced any physical activity and showed prevalence of UI as first or secondary outcomes. Exclusion criteria was defined as studies with pregnant women and patients up to 1 year postpartum. 
The gradation of impact intensity used in this analysis was used in previous publications [2, 3] which included ground reaction in body composition to determine the impact (high, moderate and low) and divided the sporting activities into one of the categories. 
The meta-analysis was performed to represent the joint prevalence of the studies and the association between prevalence values and standard error. The measures of heterogeneity of the studies was calculated using Cochran's Q test (chi-square test) and I2 (percentage of variation), indicating variation percentage among the studies. Only studies that described the number of symptomatic athletes for UI in each sport individually were analyzed.

The search strategy included 471 studies, 40 of which followed the methodological criteria for analysis. We identified 9902 women, of whom 8450 were physical activity practitioners, between 12 and 71 years of age. A total of 26 sports modalities were analyzed. The prevalence of UI ranged from 0% in low or no impact activities (skating and golf) to 80% in trampolinists. There was high heterogeneity among the included studies, and the prevalence of UI for high-impact sports was 47.52% (95% confidence interval [CI] 36.17-58.86). For medium and low impact UI, the prevalence was 33.41% (95% CI 21.21-545.62) and 20.28% (95% CI 8.48-32.07), respectively. (Figure 1) Additionally, in studies with a control group, there was an increased risk of developing UI in athletes (OR 2.83) (Figure 2).
Comparing this results with the previous systematic review, general prevalence of UI in sports was relatively similar (33,69% to 30,47%) with larger samples; however, meta-analysis presented a drop in the prevalence in high impact activities (58,10 to 47,52). This reduce is mainly due to selection of athletes (including the aspects of being a professional or recreational). An increase in the prevalence in low impact sports was revealed (12,48 to 20,28) and, in a lesser extent, in moderate impact sports (30,46 to 33,41). In both graduations the amount of athletes was not significant higher.

This is the first systematic review with meta-analysis that graduate the prevalence of UI according to different intensities of ground impact in sports. In our review the prevalence of UI increases with the ground impact. However, studies reveal three times more participants in high impact than low impact activities. The groups in high impact sports studies are more heterogeneous, with the prevalence higher in samples with greater IC. Thus, there is no comparison of strength, morphology, or muscle response between activities considered to have a distinct impact.
The definition of who is considered an athlete and the distinction of women with professional training from recreational training to analyze reported symptoms should be performed. The high heterogeneity of the studies resulted in difficulty in separating the groups in a homogeneous way, which included many studies of patients with previous pelvic surgery, multiparous and nulliparous in the same population, and the equalization of the risk of UI in women with different training time and frequency. Other factors for incontinence symptoms, such as whether the onset of incontinence was influenced by training intensity and frequency, could not be evaluated in this systematic review.

We identified an increased risk for UI in athletes in relation to the general population. Although compared to controls the prevalence is higher in women who exercise, the meta-analysis revealed that intensity of impact may not be as prejudicial as we thought. Studies with homogeneous participants are necessary to establish a relationship between the different impact intensities.

De Mattos Lourenço TR, Matsuoka PK, Baracat EC, Haddad JM. Urinary incontinence in female athletes: a systematic review. Int Urogynecol J. 2018 Dec;29(12):1757-1763
Groothausen J, Siemer H, Kemper HCG, Twisk J, Welten D. Influence of peak strain on lumbar bone mineral density: an analysis of 15-year physical activity in young males and females. Pediatr Exerc Sci 1997; 9:159-173
Torstveit MK, Sundgot-Borgen. Low bone mineral density is two to three times more prevalent in non-athletic premenopausal women than in elite athletes: a comprehensive controlled study. Br J Sports Med 2005; 39(5):282-287