Meta-analysis of pelvic floor muscle training: randomized controlled trials in incontinent women

Although many studies have reported the effectiveness of pelvic floor muscle training (PFMT) for treating female urinary incontinence, the magnitude of the effect and the optimal configuration of the parameters of the training have not been clearly determined.

The aims of this meta-analysis were: (a) to calculate the effect size of pelvic floor muscle training compared to no treatment on incontinent episodes, urine leakage amount, and perceived severity of urine loss, and (b) to identify parameters of PFMT and subjects' characteristics influencing the magnitude of the effects.

The search for relevant literature published from 1980 to 2005 consisted of using several computerized databases, citation searching, and footnote chasing. Twelve studies met the inclusion criteria, and were reviewed and coded.

The overall mean weighted effect size on incontinent episodes, urine leakage amount, and perceived severity were -0.68 (Z = 5.89, p < .001), -1.48 (Z = 2.64, p = .008), and -1.66 (Z = 1.68, p = .092), respectively. The studies with women having stress urinary incontinence showed a mean weighted effect size of -0.77 (Z = 7.03, p < .001), whereas studies with women having any type of urinary incontinence showed a mean weighted effect size of -0.47 (Z = 4.40, p < .001). The mean weighted effect size for studies including subjects over 60 years mean age was -0.54 (Z = 6.21, p < .001), whereas that of studies in which the average age was younger than 60 years was -0.94 (Z = 6.58, p < .001).

The treatment effect of PFMT on the incontinent episodes may be greater in younger women with only stress urinary incontinence. It appears that the number of daily contractions and the length of training period are not related to effect sizes on the condition that training includes at least daily 24 contractions and keeps for at least 6 weeks.