This study was a prospective cohort controlled trial conducted at the China Rehabilitation Research Center. From January 2023 to June 2023, Registered patients experiencing PPI were recruited from both outpatient and ward settings. They were all patients who had been diagnosed with PPI, either as outpatients or in the hospital. The study was approved by the ethics committee (2023–041-01), and it was conducted independently and thoroughly. Additionally, all patients were informed of the treatment process, they all signed a consent form for rehabilitation.

Inclusion criteria: Participants aged 60 to 80 years and within 1 year of surgery. Retropubic radical prostatectomy and laparoscopic radical prostatectomy. The patient's condition was stable and the wound healed well after surgery. 1 h pad test greater than 2 g [22]. Exclusion criteria were participants who withdrew or had incomplete data, had a history of other urologic procedures, had a history of neurologic disease, or had psychiatric symptoms. Initially, we planned to enroll 60 participants, 20 in each trial group, G*power software was used to calculate the sample size,to account for an alpha of 0.05, 80% power, and a between-group effect size of 60% [23]. A total of 49 patients were selected to participate in the study. Considering the different BMI [BMI = Height(m)/weight (kg)2] of the participants could affect the urinary control treatment of pelvic floor exercise [23]. We divided the patients into three groups according to the BMI range of the participants, there were 17 participants each in groups A (BMI ≤ 25) and B(26 ≤ BMI ≤ 30), and 15 participants in group C(BMI ≥ 31).

Through the model of group discussion, we developed a pelvic floor training program. Before the training, the pelvic floor anatomy and related knowledge were introduced to all patients, and the intake of caffeine-containing beverages was reduced. The treatment protocol was identical for all three groups of patients, who were all instructed in pelvic floor training by the same therapist. The training program was conducted by Pilates combined with pelvic floor muscle training. Pilates training we refer to studies developed by others [24, 25] (Table 1). The pelvic floor muscle training method was formulated according to other studies, each time the anus was contracted for 5 s, and the relaxation was 3 s. Our patients were encouraged to contract the anus as much as possible, with each contraction ensuring movement of the penis or scrotum. We let our patients experience which muscles force when flow is interrupted during voiding, and we do our best to strengthen the voluntary contraction of these muscles [26]. the PERFECT evaluation model was used for training assess [27]. All patients were required to train daily in the hospital, each training session lasted 45 min, and training lasted for 2 months. Urinary continence was evaluated every Saturday and Sunday, and data were recorded. Data was collected by nurses who were unaware of the study grouping and training methods.

The primary outcome was the 1-h pad test. The patient was asked to wear the urine pad, Participants drank as much as 500 ml of water as possible and perform activities to stimulate urine leakage, such as going up and down stairs and walking, after 1 h, the urine pad was removed and weighed, and the urine leakage was judged by calculating the weight difference of the urine pad [28]. Data from 1-h pad tests were averaged over two consecutive days. According to the difference in pad weight, the patients were divided into: no incontinence < 2 g(no UI); mild urinary incontinence 2–9.9 g(mild UI); moderate UI 10–49.9 g(moderate UI) and severe UI > 50 g (severe UI) [28]. Secondary outcomes were number of incontinent episodes [26], The modified Oxford Grading Scale for recovery of pelvic floor muscle strength [29], and International Consultation Incontinence Questionnaire (ICIQ-SF) for subjective participant recovery [30]. The data were collected at the end of every week, and the changes of the evaluation indexes were recorded by the curve graph.

SPSS 20 software was used for statistical analysis. The basic information of the patients was analyzed by one-way ANOVA test and Fisher's exact test. The 1-h pad test and the number of incontinence episodes were expressed as means and standard deviations, and the differences between groups were determined by one-way ANOVA, and the differences among the three groups and before, during and after treatment within the group were determined by 3 × 3 binary ANOVA. The Student–Newman–Keuls [S–N-K (S)] test was used for post hoc analysis of variance. The Oxford Rating Scale and the International Incontinence Consultation Questionnaire were expressed as medians and quartiles, and the Mann–Whitney rank sum test was used to determine differences between groups, and the signed rank sum test was used to determine differences before and after treatment. Spearman rank correlation was used to determine the correlation between different BMI groups and 1-h pad test, incontinence episodes, ICIQ-SF and Oxford Rating scale results after 8 weeks of treatment. P < 0.05 was considered statistically significant.