SEARCH

SEARCH BY CITATION

Keywords:

  • physical activity;
  • open and robot-assisted radical prostatectomy;
  • occupation;
  • sports;
  • household;
  • predictive factors

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References

Objective

  • To investigate the progression of all aspects (total, occupational, sports, household) of physical activity (PA) over time after radical prostatectomy (RP) and to find predictive factors for a decrease in PA.

Patients and Methods

  • In all, 240 men planned for open or robot-assisted RP were analysed.
  • All patients completed the Flemish Physical Activity Computerised Questionnaire before RP concerning PA over the past year and at 6 weeks, 3, 6 and 12 months after RP for the PA of the past month.
  • A linear model for repeated measures was used to evaluate the progression of continuous variables over time and the effect of various predictors for the progression of patients over time. A logistic regression model for repeated measures was used to evaluate binary measures.

Results

  • Total, occupational, sports and household PA levels were significantly decreased at 6 weeks after RP, but recovered quickly to approximately baseline levels from that time.
  • Predictive factors for decreased PA levels at 6 weeks after RP were a younger age (total PA level), being unskilled/semi-skilled (occupational PA level) and being unemployed/retired (household PA level).
  • RP type (open vs robot-assisted) did not influence the different PA levels at 6 weeks, 3, 6 or 12 months after RP.
  • The severity of first day incontinence and urine loss measured at 6 weeks and 3, 6 and 12 months after RP significantly affected total and/or household PA level at all time-points.

Conclusions

  • This is the first study to investigate the progression of all aspects of PA (total, occupational, sports and household) after RP and to find predictive factors for a decrease in PA.
  • All PA levels were significantly decreased at 6 weeks after RP and recovered quickly to approximately baseline levels from that time.
  • Patients that had robot-assisted RP did not have a faster recovery of PA than those that had open RP. Severity of first day incontinence and urine loss measured at 6 weeks and 3, 6 and 12 months after RP were significantly related to total and/or household PA level at all time-points.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References

For years localised or locally advanced prostate cancer have been indications for a radical prostatectomy (RP) [1-4]. Open radical prostatectomy (ORP) is performed through an open incision, and more recently robot-assisted laparoscopic RP (RALP) has been developed [4-6]. The impact of diagnosis and treatment of several cancers on the physical activity (PA) of the patient has been explored in numerous studies [7-11]. PA concerns total PA, time to return to sports, work, household function and leisure time [12]. Most studies have investigated only one aspect of PA during the first year after RP. Three studies investigated occupational PA, but patients were not assessed before RP and at fixed intervals after RP up to 12 months [6, 13, 14]. However, many studies have evaluated the progression in health-related quality of life (HRQL) [3, 15-21].

Patient- and treatment-related factors can be associated with decreased PA after surgery. Younger age, higher educational level, low physical workload, nerve-sparing status, high monthly salary and having a partner have been identified as factors with a favourable impact on postoperative PA, HRQL and return to work [6, 13, 22, 23]. Only one study has investigated the impact of RALP on the physical component scores (SF-12) of the patient [5]. They showed that the scores in the RALP group were significantly higher than those in the ORP cohort [5]. Furthermore, Hohwü et al. [6] reported that patients that had a RALP returned to work significantly faster than those that had an ORP.

Finally urinary incontinence remains a bothersome consequence of RP with an important impact on PA and HRQL [1, 24-26].

The aims of the present study were to determine the amount and duration of the decrease in the different PA levels after RP and to find predictive factors for a decrease in PA after RP.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References

This study is a longitudinal cohort study and received ethical approval from the commission medical ethics of the University Hospitals Leuven responsible for human/animal experimentation (ML5471).

In all, 278 consecutive patients that were planned for ORP or RALP for localised or locally advanced prostate cancer in the University Hospitals Leuven, between September 2009 and July 2011 were eligible. In all, 240 men were included and signed informed consent. Men with cognitive problems, non-Dutch speaking patients, patients with neurological disease (e.g. Parkinson's disease) or patients, who refused to participate, were excluded.

Patients were asked to participate when RP was planned. The surgical approach was based on the decision of the surgeon. In general, low-risk patients more often had RALP, while all high-risk patients (≥ cT3, PSA level ≥20 ng/mL and Gleason score ≥8) underwent ORP because an extended lymph node dissection was needed. Three surgeons each specialised in ORP and/or RALP completed all the operations. The first surgeon had performed >3000 ORPs, the second had performed ≈700 ORPs and 50 RALPs, and the third had performed ≈150 RALPs at the start of the study. After RP, all patients had a urinary catheter in situ for ≈12 days. In all, 173 (72%) patients were willing to register urine loss per 24 h daily until total continence was achieved. Further continence data of this group were published previously [27].

All patients were asked to complete the Flemish Physical Activity Computerized Questionnaire (FPACQ) before RP concerning their physical activities over the past year. Subsequently at 6 weeks, 3, 6 and 12 months after RP, patients reported their PA over the past month. Patients were contacted by telephone if no response was received. After RP men were counselled that they should avoid cycling and lifting during the first 6 weeks after RP.

The FPACQ

The FPACQ is a reliable and valid questionnaire [12]. The first part collects patient-related data. The second part concerns occupational activities: occupational status, working hours per week, job intensity and transport to the job. The third part collects information on sporting activities: the three most frequently performed sports, frequency and duration of each sport. The fourth part records household activities (light, moderate and vigorous). Transport during leisure time and sedentary activities were also noted.

The four different PA variables were calculated: total PA level and occupational, sports and household PA levels. All metabolic equivalent of task (MET) values used for calculating the PA variables were determined using the Ainsworth compendium of activities [28]. Using the definition for a MET as the ratio of work metabolic rate to a standard resting metabolic rate of 1.0 kcal/kg/h, one MET is considered as the resting metabolic rate during quiet sitting. Calculation of each variable is explained with an example.

The total PA level is the sum of occupational, sports, household PA levels (explained below) and active transport in leisure time (1.0 h/week × 4 MET), eating (8.8 h/week × 1.8 MET), sleeping (49 h/week × 0.9 MET) and quiet leisure time (47.7 h/week × 1.5 MET) = 307 MET-h/week.

A patient worked 38 h/week with 20% light, 70% moderate and 10% vigorous activities. The occupational PA level was: (38 h/week × 20% × 2 MET) + (38 h/week × 70% × 3 MET) + (38 h/week × 10% × 4 MET) = 110 MET-h/week. He drove 1.3 h/week for work by car. Consequently, the PA level for transport for work was: 1.3 h/week × 1.5 MET = 2 MET-h/week.

This patient performed two sports. He danced 2 h/week and walked 0.7 h/week. The sports PA level was: (2 h/week × 5.5 MET) + (0.7 h/week × 3.5 MET) = 13 MET-h/week. He also performed 10 h/week light, 4 h/week moderate and 1.5 h/week vigorous household activities. The household PA level was: (10 h/week × 2.5 MET) + (4 h/week × 3.5 MET) + (1.5 h/week × 4.5 MET) = 46 MET-h/week.

Predictive Factors

Patient- and treatment-related factors were prospectively collected. Patient-related factors were age, body mass index (BMI), partner status, educational level, employment status (retired, employed [labourer/self-employed], unemployed) and smoking behaviour. Additionally, incontinence on the first day after catheter removal and urine loss at 6 weeks and 3, 6 and 12 months after RP (urine loss = mean urine loss of 7 days around the time the PA level was measured) were assessed. These factors were collected with the FPACQ, except for incontinence (criterion >0 g of urine loss per day [24-h pad test]). Treatment-related factors included type of RP and nerve-sparing status and were obtained from the surgeon.

Statistical Analysis

A linear model for repeated measures was used to evaluate the progression of continuous variables (total, occupational, sports and household PA levels) over time. The model used an unstructured covariance matrix for measurements over time and has, compared with the classical repeated measures anova, the advantage that subjects with missing measurements were still included in the analysis and that the results are still valid when drop-out is missing at random [29]. The same models were used to analyse the effect of various predictors for the progression of patients over time. Differences in progression over time according to patients' characteristics were evaluated by testing time by predictor interaction effects. A logistic regression model for repeated measures, with unstructured covariance matrix, was used to evaluate longitudinal binary measures, e.g. job status or practicing sports. The analysis of the occupational PA level was restricted to preoperatively employed patients and the sports PA level to patients practicing some sports preoperatively. Patients, who retired during follow-up, were only taken into account at the time points they were still employed. Tukey–Kramer or Holm adjustment was used for multiple post hoc comparisons.

Results should be interpreted with care given the multitude of models and P values.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References

In all, 240 patients completed the FPACQ before RP, and 232 (97%), 231 (96%), 225 (94%) and 224 (93%) at 6 weeks, 3, 6 and 12 months after RP, respectively. Baseline characteristics for the whole group are shown in Table 1.

Table 1. Baseline characteristics (N = 240) and incontinence data (N = 173)
VariableValue
Mean (sd) age, years61.7 (6.3)
N (%): 
BMI, kg/m2: 
≤2578 (32.5)
25.1–30.0135 (56.3)
>30.027 (11.2)
Employment status: 
Retired135 (56.2)
Employed89 (37.1)
Unemployed16 (6.7)
Smoking status: 
Non-smoker204 (85.0)
Smoker36 (15.0)
Partner: 
No13 (5.4)
Yes227 (94.6)
Educational level: 
Semi-/unskilled110 (46.0)
Highly skilled129 (54.0)
Type of RP: 
Open174 (72.5)
Robot-assisted66 (27.5)
Surgical margins status: 
Negative175 (73.5)
Positive/doubtful63 (26.5)
Adjuvant radiotherapy 
No208 (86.7)
Yes32 (13.3)
Missing2 (0.9)
Median (range) incontinence day 1 after catheter removal, g132 (0.0–2648.0)
N (%): 
Incidence of incontinence: 
Preoperatively53 (32.3)
At 1 month85 (49.1)
At 3 months35 (20.3)
At 6 months11 (6.5)
At 12 months6 (3.6)

Progression of PA (Table 2)

Table 2. Progression of PA levels over time
PAPreoperative6 Weeks3 Months6 Months12 Months
  1. P value of pairwise differences between preoperative and the particular point in time.

Total PA:     
Mean estimate, MET-h/week268.70240.81260.00263.24263.57
95% CI264.82–272.58237.97–243.64256.18–263.81259.16–267.31259.71–267.43
P <0.001<0.0010.0160.013
Occupational PA:     
Mean estimate, MET-h/week107.0746.9683.9797.6096.63
95% CI99.92–114.2135.30–58.6373.61–94.3288.11–107.0887.79–105.47
P <0.001<0.0010.0790.016
Sports PA:     
Mean estimate, MET-h/week32.939.1724.5225.8927.21
95% CI28.91–36.956.16–12.1920.11–28.9221.29–30.5022.48–31.94
P <0.0010.0010.0040.051
Household PA:     
Mean estimate, MET-h/week37.4723.8836.8735.7435.30
95% CI33.48–41.4720.46–27.2932.40–41.3431.33–40.1531.30–39.30
P <0.0010.9990.9210.790

The total PA level decreased by 10% at 6 weeks after RP (from 268.7 to 240.81 MET-h/week). At 3, 6 and 12 months after RP, total PA levels were respectively, 3%, 2% and 2% lower than before RP (Fig. 1).

figure

Figure 1. Progression of the total PA levels of patients with prostate cancer from before to 12 months after RP.

Download figure to PowerPoint

A similar trend was found for occupational, sports and household PA levels (Fig. 2). For patients employed before RP (89 patients), the average amount of MET-h/week spent on occupational activities decreased by 56% at 6 weeks after RP. At 3, 6 and 12 months after RP, occupational activities were respectively decreased by 22%, 9% and 10%. Before RP, 89 (85%) of 105 non-retired patients were employed. At 6 weeks, 3, 6 and 12 months 42%, 70%, 81% and 80% of patients employed before RP had resumed their preoperative occupational activities. The sports PA level of sport-active patients before RP (171 patients) was reduced with 72% at 6 weeks after RP. At 3, 6 and 12 months after RP sports PA level was still respectively 26%, 21% and 17% lower than before RP. Only 42%, 73%, 73% and 74% of sport-active patients before RP were practicing some sports between 1 and 12 months after RP. For patients not performing a sport before RP, 20% practiced a sport at 12 months after RP.

figure

Figure 2. Progression of the occupational, sports and household PA levels of patients with prostate cancer from before to 12 months after RP.

Download figure to PowerPoint

The household PA level decreased by 36% in the first month after surgery (P < 0.01). Afterwards, the household PA level recovered approximately to baseline in all 240 patients.

Predictive Factors for Progression of PA Levels (Table 3, Fig. 3)

figure

Figure 3. Predictive factors for the progression of the (A) total, (B) occupational and (C) household PA levels from before to 12 months after RP.

Download figure to PowerPoint

Table 3. Predictive factors for the progression of PA levels
 AgeEducational levelEmployment status
Employed vs unemployedEmployed vs retired
  1. A positive value for the difference in the MET-h/week estimate at 6 weeks indicates a stronger decrease in the patients aged <60 years/high education/employed group, a negative value for the difference in the MET-h/week estimate at 6 weeks indicates a stronger decrease in the patients aged ≥60 years/low education/unemployed or retired groups.

Total PA:    
Difference in MET-h/week estimate at 6 weeks9.99   
P0.037   
CI2.49 to 17.49   
Occupational PA:    
Difference in MET-h/week estimate at 6 weeks –32.06  
P 0.048  
CI –56.89 to –7.23  
Household PA:    
Difference in MET-h/week estimate at 6 weeks  –23.17–15.29
P  0.0200.011
CI  –38.29 to –8.05–24.62 to –5.97

Younger patients had a greater decrease in total PA level than older patients at 6 weeks after RP, because younger patients had a higher PA level than older patients at all time-points, except at 6 weeks (Fig. 3). Additionally, highly active patients before RP had a larger decrease compared with those with low PA levels. A greater decrease in occupational activities was associated with being unskilled/semi-skilled at 6 weeks after RP (Table 3, Fig. 3). Before RP, 25% of working patients were self-employed. At 6 weeks, 3, 6 and 12 months after RP 85%, 100%, 100% and 100% of the self-employed patients were working again vs only 38%, 77%, 88% and 92% of the employees. Seven patients, all employees, retired in the first postoperative year. No predictive factor could be found for a greater decrease in sports PA level. The only predictive factor for a decrease in the household PA level at 6 weeks after RP was being unemployed or retired (Table 3, Fig. 3).

There was no interaction effect between RP type (ORP/RALP) and time, indicating similar progression over time for both groups (Table 4). Furthermore the PA level at almost all time-points was higher in the RALP group, but the difference was only significant for total PA (P = 0.02). After correction for the significant patient- and treatment-related characteristics (age, risk group and nerve-sparing) the difference in total PA between both groups clearly diminished and statistical significance disappeared (P = 0.37; data not shown). For sports, occupational and household PA levels, differences were not significantly different between ORP and RALP (Fig. 4).

figure

Figure 4. Progression of the (A) total, (B) occupational, (C) sports and (D) household PA levels of patients with prostate cancer from before to 12 months after RP.

Download figure to PowerPoint

Table 4. Main effects of urinary incontinence and type of RP.
 Type of RP (ORP/RALP)Urine loss at day 1 after catheter removal (per 10 g increase in urine loss)Urine loss as continuous longitudinal measurement (per 10 g increase in urine loss)
Total PA:   
Interaction effect, P0.4590.9210.681
Main effect–6.31–0.07–0.11
Main effect, P0.0260.0100.110
CI–11.85 to –0.77–0.13 to 0.02–0.25 to 0.02
Occupational PA:   
Interaction effect, P0.1150.8800.697
Main effect1.77–0.21–0.70
Main effect, P0.7930.0620.079
CI11.63 to 15.18–0.44 to 0.01–1.48 to 0.08
Sports PA:   
Interaction effect, P1.5430.7860.186
Main effect–1.740.010.09
Main effect, P0.5250.7560.142
CI–7.15 to 3.65–0.04 to 0.05–0.03 to 0.22
Household PA:   
Interaction effect, P0.9710.3530.330
Main effect–3.17–0.07–0.20
Main effect, P0.3320.030.013
CI–9.60 to 3.25–0.14 to –0.01–0.36 to –0.04

The effect of first day incontinence and urine loss (measured at 6 weeks and 3, 6, 12 months after RP) on total, occupational, sports and household PA levels was not different at the various time points (P > 0.05). However, the amount of first day urine loss had a significant effect on the absolute level of total and household PA. For every 10 g increase in first day incontinence a patient was, 0.07 MET-h/week (total, P = 0.01; household, P = 0.03) less physically active. Additionally, for every 10 g increase in urine loss at 6 weeks, 3, 6 and 12 months, a patient was overall 0.2 MET-h/week less physically active for the household PA level (P = 0.01).

Patient's BMI, nerve-sparing status, having a partner, smoking behaviour, surgical margins status and adjuvant radiotherapy during the first year after RP did not predict a different progression of PA over time.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References

In the present study the total PA level of patients after RP decreased by 10% at 6 weeks after RP. Additionally, we found that patients recovered to approximately baseline PA levels from 3 months after RP. Previous studies have only been concerned with HRQL and have found that at 3 months after RP, 60% of patients reached baseline HRQL levels [18, 20]. The occupational PA levels of the present patients decreased by 56%, 22%, 9% and 10%, respectively, at 6 weeks, 3, 6 and 12 months after RP. No other study was found that examined postoperative occupational activities as part of PA and expressed in MET-h/week. Other studies have investigated return to work after RP. A median of 25–49 days of sick leave was found after ORP [6, 14] and 11 days of sick leave after RALP [6]. Also, 27.9% and 72.4% of the patients after respectively ORP and RALP returned to work 6 weeks after RP [6]. In the present study both groups had 6 weeks of paid sick leave prescribed.

The sports PA level in the present patients decreased by 72% in the first month after RP. A partial explanation is that patients were counselled not to cycle and not to lift in the first 6 weeks after RP. The household PA levels were only significantly different from preoperative levels up to 6 weeks after RP. We could not find any other studies evaluating sport participation or household PA levels after RP to compare with.

Previously published studies on PA levels after RP have some limitations. Most studies investigated only one aspect of PA, e.g. return to work [6, 13] or HRQL [15, 17, 18, 20, 30]. The postoperative follow-up period was variable and often rather short [5, 13, 15, 26]. Most studies only investigated one surgical approach [13, 16, 26]. One study examined a small sample concerning lifetime PA before RP, but compared this with postoperative HRQL [15].

In the present study, being younger was a predictive factor for a decreased total PA level at 6 weeks after RP. Similar to the present study, Sultan et al. [13] found being younger was associated with a slower return to work. Bradley et al. [22], on the contrary, concluded that older men were less likely to be employed 6 months after RP. In the present study preoperatively more active patients had a larger decrease in PA at all-time points. According to Mina et al. [15] higher levels of total PA before RP were associated with a lessened HRQL reduction at 4 weeks after RP.

Furthermore, educational level predicted a decreased occupational PA level at 6 weeks after RP in the present study. The importance of type of profession on time to work resumption has been confirmed in other studies [6, 13]. Also, we found that being unemployed/retired was a predictive factor for a decrease in household PA levels at 6 weeks after RP. No other studies have evaluated household PA levels. According to Sultan et al. [13] married men returned faster to full-time work than unmarried men. This was not confirmed in the present study.

Additionally RP type did not influence the different PA levels at 6 weeks, 3, 6 or 12 months after RP. On the contrary, Miller et al. [5] reported a faster return to baseline physical component scores after RALP.

Two studies found a significant effect of incontinence on PA [16, 26]. In the present study, patients with a large amount of first day incontinence were significantly less physically active for total and household PA levels than patients with a smaller amount of first day incontinence. Similarly, patients with a larger amount of urine loss (longitudinally measured) at 6 weeks, 3, 6 and 12 months were significantly less active in household tasks than patients with a small amount of urine loss.

The present study has several strengths. Patients were measured prospectively before RP and at fixed time intervals after RP up to 12 months. All aspects of PA were measured: occupational, sports and household PAs, transport during leisure time and sedentary activities. The return rate of the questionnaires was very high. Furthermore, this study is unique in finding a determinant of return to PA in incontinence. Finally, patients were operated on by three experienced surgeons, who used highly standardised procedures. However, due to the fact that only three surgeons provided the surgery and a strong relationship existed between surgical expertise and surgeon, we were not able to correct for surgical experience.

The limitations of the present study were that the FPACQ was tested for reliability and validity in healthy adults and not specifically in patients with prostate cancer. As sick leave and avoidance of PA were prescribed for 6 weeks, earlier return to PA for RALP in this first period may have been masked. Only 30% of the cohort underwent RALP. Finally not all, but still 72% of patients were willing to record daily urine loss until total continence was achieved.

In conclusion, this is the first study to investigate the progression of all aspects of PA (total, occupational, sports and household) after RP, to find predictive factors for a decrease in PA and to assess the impact of type of RP and incontinence on PA. All aspects of PA decreased significantly at 6 weeks after RP. Predictive factors for decreased PA levels at 6 weeks after RP were a younger age (total PA level), being unskilled/semi-skilled (occupational PA level) and being unemployed/retired (household PA level). The RP type did not influence the different PA levels at 6 weeks, 3, 6 or 12 months after RP. The severity of first day incontinence had a significant effect on total and household PA levels. Urine loss measured at 6 weeks and 3, 6 and 12 months after RP significantly affected the household PA level.

Funding

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References

This study was funded by the Agency for Innovation by Science and Technology (IWT). The IWT had no role in study design and realisation, data collection, analysis and interpretation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. Funding
  9. References
Abbreviations
BMI

body mass index

FPACQ

Flemish Physical Activity Computerized Questionnaire

HRQL

health-related quality of life

MET

metabolic equivalent of task

PA

physical activity

(O)RP

(open) radical prostatectomy

RALP

robot-assisted laparoscopic RP