Association of physical activity with overall mortality among long‐term testicular cancer survivors: A longitudinal study

Physical activity (PA) has been associated with reduced mortality among cancer survivors, but no study has focused on testicular cancer survivors (TCSs). We aimed to investigate the association of PA measured twice during survivorship with overall mortality in TCSs. TCSs treated during 1980 to 1994 participated in a nationwide longitudinal survey between 1998 to 2002 (S1: n = 1392) and 2007 to 2009 (S2: n = 1011). PA was self‐reported by asking for the average hours per week of leisure‐time PA in the past year. Responses were converted into metabolic equivalent task hours/week (MET‐h/wk) and participants were categorized into: Inactives (0 MET‐h/wk), Low‐Actives (2‐6 MET‐h/wk), Actives (10‐18 MET‐h/wk) and High‐Actives (20‐48 MET‐h/wk). Mortality from S1 and S2, respectively, was analyzed using the Kaplan‐Meier estimator and Cox proportional hazards models until the End of Study (December 31, 2020). Mean age at S1 was 45 years (SD 10.2). Nineteen percent (n = 268) of TCSs died between S1 and EoS, with 138 dying after S2. Compared to Inactives at S1, the mortality risk among Actives was 51% lower (HR 0.49, 95% CI: 0.29‐0.84) with no further mortality reduction among High‐Actives. At S2, the mortality risk was at least 60% lower among the Actives, High‐Actives and even the Low‐Actives compared to the Inactives. Persistent Actives (≥10 MET‐h/wk at S1 and S2) had a 51% lower mortality risk compared to Persistent Inactives (<10 MET‐h/wk at S1 and S2; HR 0.49, 95% CI: 0.30‐0.82). During long‐term survivorship after TC treatment, regular and maintained PA were associated with an overall mortality risk reduction of at least 50%.


What's new?
Testicular cancer survivors have increased risk of cardiovascular disease, second cancers and premature death. Regular physical activity has been shown to reduce overall mortality among cancer survivors. This longitudinal nationwide survey is the first study to investigate the association between physical activity and overall mortality among testicular cancer survivors. Compared to physically inactive testicular cancer survivors, physically active testicular cancer survivors treated between 1980 and 1994 had a reduced overall mortality risk of 50%. Clinicians and testicular cancer survivors should be aware of the association between physical activity and reduced risk of mortality during testicular cancer survivorship.

| INTRODUCTION
Testicular cancer (TC) is the most common malignancy among men during adolescence and young adulthood. With a median age of 36 years at diagnosis and a current 15-year relative survival rate of more than 98%, 1 TC survivors (TCSs) have long life-expectancy.
Studies, however, indicate that modern TC treatment increases the risk of cardiovascular disease, second cancer and premature mortality. [2][3][4][5][6][7] While physical activity (PA) has been associated with reduced risk of recurrence and improved survival after breast, prostate and colorectal cancer, 8,9 no studies have focused on PA and mortality in TCSs. 10 Moreover, it is unknown whether increasing levels of PA during TC survivorship reduce the risk of death among TCSs, or how change in PA during TC survivorship is associated with mortality.
The Norwegian Testicular Cancer Project (NTCP) 11 is a nation-wide longitudinal study, which has collected data among long-term TCSs on adverse health outcomes, health-related quality of life and lifestyle.
The NTCP provides the opportunity to explore the association between levels of PA and overall mortality in TCSs. The primary aim of our study was to investigate if rising levels of PA, measured at two different time-points during TC survivorship, predicts decreasing risk of overall mortality. The secondary aim was to explore the association between change in PA during survivorship and overall mortality. 2009, provide data for the current study. 11 The study population is described in detail previously. 2,12 Dependent on histology and stage, postorchiectomy treatment consisted of no further therapy or retroperitoneal lymph node dissection (SURGERY-group), abdominal radiotherapy only (RAD-group), or platinum-based chemotherapy with or without major surgery or radiotherapy (CHEMO-group).  Figure S1). The descriptor "not sweaty/breathless" was considered to describe light to moderate PA, corresponding to a metabolic equivalent task (MET)-value of 4, whereas "sweaty/breathless" described vigorous PA, corresponding to a MET-value of 8. We assigned the time spent on PA for the <1 h/wk category to be 0.5 hours, for the 1 to 2 hours to be 1.5 hours and for the ≥3 hours to be 4 hours. For the analysis of PA change from S1 to S2, four categories were defined ( Figure S2): Persistent Actives (Overall Actives at S1 and S2), Improvers (Overall Inactives at S1 but Overall Actives at S2), Decliners (Overall Actives at S1 but Overall Inactives at S2) and Persistent Inactives (Overall Inactives at S1 and S2).

| Physical activity assessments
T A B L E 1 Socio-demographic, cancer-related, comorbidity and lifestyle variables by four physical activity groups at Survey 1, and total physical activity at Survey 1 and Survey 2.

| Outcome
The outcome of the study was overall mortality from S1 and S2, respectively, to End of Study (EoS: December 31, 2020). For each TCS, the CRN provided the date of last observation, that is, death, emigration or EoS, whichever occurred first. The postsurvey observation times ranged from the date of S1 or S2 until date of last observation.

| Other covariates
The multivariable models were adjusted for confounding factors including age at S1 or S2, level of education (high [college or university] vs low [high-school or less]); self-reported major somatic comorbidity (≥1 of the following myocardial infarction, angina pectoris, stroke, diabetes, history of non-TC cancer; yes vs no); current daily smoking (yes vs no); hazardous alcohol use (yes vs no) 14 ; and treatment groups (SURGERY-group vs RAD-group vs CHEMO-group).
Due to the well-known correlation between treatment and histology and initial stage, treatment was the only independent variable included in the multivariable models.

| Statistics
Descriptive statistics were presented by means and SDs for continuous variables, and by absolute numbers and percentages for categorical variables. Using the Kaplan-Meier estimator, crude overall mortality rates were studied for the four PA groups after S1 and S2 respectively, and after S2 the mortality rates were studied for PA change. The log-rank test was used for assessment of differences between groups. Three multivariable Cox proportional hazards regression models were applied to estimate the association between overall mortality and levels of PA at S1, levels of PA at S2 and PA changes T A B L E 2 Association between physical activity at Survey 1 and Survey 2 and overall mortality in long term testicular cancer survivors. Overall mortality model is adjusted for age at S1, education, treatment, major somatic co-morbidity, smoking and hazardous alcohol use. b Overall mortality model is adjusted for education, treatment and hazardous alcohol use at S1; and age, major somatic co-morbidity and smoking at S2. from S1 to S2, adjusted for other covariates. For the multivariable adjusted analyses only HR's for PA groups should be interpreted.
Other HR's are only included for comparison with unadjusted HR's.
The proportional hazard assumptions were not violated as tested by the estat phtest command in Stata version 17.0. The other statistical analyses were performed using SPSS for Windows, version 26.

| RESULTS
TCSs: At S1, 1437 TCSs completed the questionnaire (response rate 79%), of whom 45 were excluded from all analyses due to missing responses to both PA intensity categories ( Figure S3). Age and cancer-related variables were similar among the 1392 S1 evaluable participants and the 376 nonresponders (Table S1). S1: On average 12 years had elapsed since orchiectomy (Table S2), and S1 participants were aged mean 45 years (SD 10.2) (  Figure S2).
PA at S1 and overall mortality: After a median post-S1 observation time of 20 years (range 0-22) (Table S2), 268 of the 1392 S1-participants (19%) had died (  Table S3). The Kaplan-Meier plot revealed no significant difference of crude mortality rates between the Inactives and the Low-Actives, but documented a significant lower mortality rate for the Actives and for the High-Actives compared to the Inactives ( Figure 1A), though without difference between the Actives and High-Actives. Results from multivariable Cox regression analysis (  Table S3). The Kaplan-Meier plot showed that each of the three active groups had lower overall mortality rates than the Inactives (P = .001), but without significant differences between the three active groups ( Figure 1B). The P P P P P P F I G U R E 1 Overall mortality and physical activity at Survey 1 (A) and Survey 2 (B).

P P P
F I G U R E 2 Overall mortality and change in physical activity from Survey 1 to Survey 2.
Cox regression analysis showed a reduction of overall mortality by at least 60% in the Low-Actives, Actives or High-Actives compared to the Inactives (Table 2). Previous treatment with radiotherapy or chemotherapy did no longer predict overall mortality at S2.    Overall mortality models are adjusted for education, treatment and hazardous alcohol use at S1; and age, major somatic co-morbidity and smoking at S2.
Friedenreich et al found a significant reduction in all-cause mortality up to 10 to 15 MET-h/wk, with less reduction associated with PA above 20 MET-h/wk. 10 In contrast, Holick et al observed a doseresponse association between multiple levels of postdiagnosis PA and overall mortality in breast cancer survivors. 18 We cannot rule out that lack of a dose-response association in our study is caused by methodological weaknesses such as an imprecise self-report of PA intensity and duration, problems recalling the actual PA level over the past year and/or a small sample. On the other hand, our observation is in line with the nonlinear association observed in a recent meta-analysis among adults in the general population, demonstrating that mortality risk decreased with increasing objectively assessed PA up to a level corresponding to the WHO recommendations, and thereafter plateaued. 19 At S2, in contrast to the results at S1, PA at the level of 2 to 6 MET-h/wk also predicted reduced mortality risk of more than 60%. At this time of assessment, the participating TCSs were almost 10 years older than at S1.