The 38 patients were aged 48 to 73 years, most were married (71.1%), without tertiary education (71.1%), and classed as moderately active or active (76.3%) based on the Godin Leisure-time Physical Activity Questionnaire (Table 1). BMI was 28.5 ± 3.4 kg.m−2 with most patients overweight or obese (86.9%). The median time since prostate cancer diagnosis was 1.0 month (IQR: 1.0 to 2.0 months), with median prostate-specific antigen (PSA) levels of 7.0 ng.ml−1 (IQR: 3.9 to 9.5 ng.ml−1) and Gleason Score of 7.0 (IQR: 7.0 to 7.0). Most patients underwent robotic-assisted laparoscopic prostatectomy (92.1%).

Three patients in prehabilitation and six from rehabilitation withdrew following baseline measurements preferring not to be in the rehabilitation group (n = 5), opting for radiation treatment (n = 1), diagnosis of brain cancer (n = 1), running injury not related to the exercise programme (n = 1), and medical complications after surgery (n = 1) (Fig. 1). Patients in prehabilitation attended 73.7% of scheduled exercise sessions while those in rehabilitation attended 93.7% of scheduled sessions. The prehabilitation group's lower attendance was attributed to patients attending medical appointments in relation to surgery. There were no exercise-related adverse events.

There were no differences between groups at baseline (Fig. 2). Across the study time points, there was no significant interaction for leg or chest press strength; however, there was a significant effect of time (p < 0.001). In the pre-surgery phase, prehabilitation significantly improved both leg (17.2 kg, p < 0.001) and chest press strength (2.9 kg, p = 0.001) with rehabilitation also experiencing a significant increase of 6.7 kg for leg press (p < 0.001). However, both groups exhibited significant reductions for chest (Prehabilitation: 4.8 kg, p < 0.001; Rehabilitation: 3.9 kg, p < 0.001) and leg press (Prehabilitation: 8.9 kg, p = 0.024; Rehabilitation: 8.7 kg, p = 0.012) strength in the early post-surgery phase. During late post-surgery phase, significant improvements of 5.0 kg (p = 0.003) and 14.6 kg (p < 0.001) were observed in leg press for both prehabilitation and rehabilitation, respectively, with only rehabilitation improving chest press strength by 6.8 kg (p < 0.001). As a result, when comparing 12 weeks post-surgery to baseline, there was no significant difference in muscle strength between prehabilitation and rehabilitation. Comparable results were observed when analysing complete cases except for leg press strength (p = 0.042) which was higher in prehabilitation at 12 weeks post-surgery compared to baseline (Table S1).

Muscle strength absolute values and change over the assessment time points. Results are presented as mean and standard error. astatistically within-group change compared to baseline; bstatistically within-group change compared to 6-weeks post-surgery; cstatistically within-group change compared to 12-weeks post-surgery

There was no difference between groups at baseline for physical function (p = 0.060–0.685). Over the study period, there were no significant interactions except for 6-m usual walk (p = 0.033) and a significant time effect for 400-m walk, chair rise, stair climb, 6-m fast and backward walk tests (p =  < 0.001–0.001) (Fig. 3). During the pre-surgery phase, there were significant reductions (improvement) for prehabilitation of –  14.9 s in the 400-m walk, –  1.3 s in chair rise, –  0.2 s and –  1.8 s in 6-m fast and backward walk tests, respectively, (p =  < 0.001–0.028). The rehabilitation group also had significant reductions of –  9.8 s in the 400-m walk, –  0.8 s in chair rise, –  0.4, –  0.2 and –  2.8 s for 6-m usual, fast, and backward walk tests, respectively, (p =  ≤ 0.001–0.012). Physical function during the early post-surgery phase (p = 0.152–1.000) was maintained through to the late post-surgery phase (p = 0.170–1.000) for prehabilitation except for a 0.7 s reduction in the chair rise test (p = 0.004). During the early post-surgery phase, rehabilitation improved with a reduced 6-m usual (–  0.3 s, p = 0.019) and backward walk time (–  1.1 s, p = 0.028), however, chair rise time increased (0.3 s, p = 0.033). The 400-m walk (–  12.0 s, p = 0.005) was the only improvement in rehabilitation during late post-surgery phase.

Physical function absolute values and change over the different assessment time points. Results are presented as mean and standard error. astatistically within-group change compared to baseline; bstatistically within-group change compared to pre-surgery; cstatistically within-group change compared to 6-weeks post-surgery

When comparing 12 weeks post-surgery to baseline, prehabilitation had significant improvement of –  16.3 s (95% CI: –  26.2 to –  6.3 s, p < 0.001) in 400-m walk, –  1.7 s (95% CI: –  3.0 to –  0.4 s, p = 0.004) in chair rise, –  0.3 s (95% CI: –  0.5 to –  0.0 s, p = 0.050) in stair climb, − 0.2 s (95% CI: –  0.4 to –  0.0 s, p = 0.047) in 6-m fast walk and –  3.9 s (95% CI: –  6.6 to –  1.3 s, p = 0.001) in backwards walk. In the rehabilitation group, there was a significant improvement of –  16.1 s (95% CI: –  24.8 to –  7.4 s, p < 0.001) in 400-m walk, –  1.2 s (95% CI: –  2.3 to –  0.2 s, p = 0.010) in chair rise, –  0.4 s (95% CI: –  0.7 to –  0.0 s, p = 0.034) in stair climb, –  0.5 s (95% CI: –  0.9 to –  0.2 s, p < 0.001) in 6-m usual walk and –  3.3 s (95% CI: –  5.6 to –  1.0 s, p = 0.001) in backwards walk following exercise post-surgery compared to baseline. As a result, there was no significant difference in physical function between groups at 12 weeks post-surgery. Similar results were observed when analysing complete cases except for a significant reduction (improvement) in chair rise (p = 0.036) for rehabilitation comparing 6 to 12 weeks post-surgery (Table S2).

At baseline, prehabilitation had lower whole-body LM (p = 0.007) and trunk FM (p = 0.038) compared to rehabilitation (Table 2). There were no significant interactions for body composition (p = 0.080–0.586) except for body fat % (p = 0.029), with a significant effect of time for LM, FM, and trunk FM (p =  < 0.001–0.041). No differences were found in either group during the pre-surgery phase; however, in early post-surgery phase, both groups lost LM (Prehabilitation 1.6 kg, p = 0.008; Rehabilitation 1.1 kg, p = 0.004) with prehabilitation having an increase of 1.7% in percent body fat (p = 0.007). There were no differences in body composition during the late post-surgery phase. When comparing 12 weeks post-surgery to baseline, there was a significant decrease in whole-body FM of 1.1 kg (95% CI: –  2.1 to –  0.2 kg, p = 0.008) and trunk FM of 0.7 kg (95% CI: –  1.3 to –  0.2 kg, p = 0.005) in rehabilitation with no significant differences for prehabilitation. When analysing complete cases, comparable results were observed for both groups (Table S3).

At baseline, there was no difference between groups for QoL and fatigue (Table 3). Over the course of the study, there were no interactions but a significant time effect for both QoL and fatigue (p < 0.001). During the pre-surgery phase, fatigue was significantly reduced (p = 0.002) for prehabilitation with no change in rehabilitation. Both groups had an increase in fatigue at 2 weeks post-surgery which was then reduced at 6 and 12 weeks post-surgery. There was no change in QoL pre-surgery; however, there was a substantial decline in both groups at 2 weeks post-surgery which then recovered to baseline levels at 12 weeks post-surgery. Similar results were observed for both groups in complete case analyses (Table S4).

At 2 weeks post-surgery, there was no difference between groups for urinary incontinence (p = 0.790) (Table 3). Across the post-surgery time points, there was no interaction (p = 0.884) but a significant time effect (p < 0.001). Rehabilitation had a significant reduction of 338.5 g (95% CI: –  615.6 to –  61.4 g) in the 24-h pad test (p = 0.010) between 2 and 12 weeks post-surgery, while prehabilitation reduction approached statistical significance (262.7 g, p = 0.067). Results were similar in sensitivity analyses (Table S4). For hospital LOS, there was no difference between groups (Prehabilitation, 2.9 ± 1.4 days vs. Rehabilitation, 2.5 ± 1.3 days; p = 0.473).