EURURO Vol. 72 No. 6

937 1030

recreational PA ( 17.5 MET-h/wk or 5 h of moderate-
intensity activity per week), including walking, was
associated with a lower risk of PCSM among men with
lower-risk tumors but not among men with high-risk
tumors. Higher postdiagnosis recreational PA was associat-
ed with a lower risk of overall PCSM, and walking after
diagnosis also had a suggestive inverse association.
To date, four cohort studies showed inverse associations
between PA and risk of PCSM or progression
however, these results should be interpreted cautiously
because reverse causation is a concern when PA is assessed
close to death date. In a Swedish cohort study, exercising
>1 h/wk and walking/bicycling >20 min/d after diagnosis
(not adjusted for other activities) were associated with a
lower risk of PCSM . After an 18-mo lag was applied, the
association of exercise was attenuated and became
nonsignificant, although the association for walking/bicy-
cling remained statistically significant. Similarly, in a
Canadian cohort study, the inverse association between
postdiagnosis recreational PA and PCSM was attenuated
and nonsignificant when excluding men with metastatic
prostate cancer or those who reported PA within 1 yr of
death . The nonsignificant results may also be due to
reduced statistical power. In the Health Professionals
Follow-up Study, engaging in >3 versus <1 h/wk of
vigorous activity after diagnosis was associated with a
61% lower risk of PCSM . Although a 4–6-yr lag was
applied in that analysis , there is still a concern that the
inverse association could partially be due to reverse
causation since men with advanced disease may decrease
their vigorous intensity activity. In an independent study,
the same research group examined PA with prostate cancer
progression, an outcome less subject to reverse causation
than mortality . In that analysis, Richman et al
found that vigorous activity had a suggestive but not
statistically significant inverse association with progres-
sion, whereas brisk walking (>3 miles/h), among men who
did not engage in vigorous activity, was inversely associated
with progression.
To reduce reverse causation bias, we excluded deaths that
occurred within 4 yr of postdiagnosis questionnaire comple-
tion, consistent with the analysis of the Health Professionals
Follow-up Study Moreover, we examined prediagnosis
recreational PA as a surrogate of long-term recreational PA to
confirm our findings. In this study, pre- and postdiagnosis
continuous recreational MET–h/wk were moderately corre-
lated (Pearson correlation coefficient = 0.4). The inverse
association of prediagnosis recreational PA with PCSM
observed among men with lower-risk tumors supports the
possibility that PAmight play a role in tumor progression and
thus the inverse association with postdiagnosis recreational
PA was not purely due to reverse causation.
An inverse association between PA and prostate cancer
progression may be due to reduced insulin resistance,
insulin-like growth factor-1, sex hormone levels, inflam-
mation, and increasing immune function among more
physically active men Recent human studies
further suggest that prediagnosis PA can influence gene
expression relating to cell cycling, DNA repair, and oxidative
stress pathways among others in normal prostate tissues
, and influence vascularization relating to progression
in prostate tumors .
Strengths of this study include its large sample size, long-
term follow-up, and assessment of both pre- and post-
diagnosis recreational PA. A limitation of this study, also
seen in most large cohort studies, is potential measurement
error from one-time self-reported PA, which can attenuate
the true associations. Although the instrument used in the
present study has not been validated in this population, it is
similar to that used and validated in the Health
[29_TD$DIFF]Professionals Follow-up Study, a prospective study with
similar participant characteristics . These measures
have also been associated with various cancers in the CPS-II
Nutrition Cohort . An additional limitation is that the
association between PA and lower risk of PCSM could partly
be due to more intensive PSA screening or medical
surveillance among physically active men, resulting in less
advanced disease at diagnosis. In our analysis, physically
active men were slightly more likely to have a self-reported
history of PSA testing, and lower stage and grade cancer at
diagnosis than less active men. Although our analysis
controlled for both a simple measure of self-reported
history of PSA testing and disease stage and grade, positive
confounding by severity of disease at diagnosis cannot be
excluded. Finally, the lack of information on treatment
completion date, adjuvant therapies, adverse effects from
treatment, and tumor recurrence limited us from better
controlling for confounding by treatment and examining
progression more directly. Residual confounding due to
other unmeasured confounders may also exist.
5. Conclusions
In conclusion, we found higher levels of postdiagnosis
recreational PA were associated with lower PCSM. Pre-
diagnosis recreational PA, including walking, was associat-
ed with a lower risk of PCSM among men with lower-risk
tumors. Given other evidence regarding the beneficial
effects of aerobic exercise on reducing fatigue, and
improving the quality of life and muscular fitness among
prostate cancer survivors our results provide
further motivation for prostate cancer survivors to adhere
to or exceed recommendations for moderate- to vigorous-
intensity activities. This study also supports the importance
of ongoing and future clinical trials to assess the influence of
PA on tumor progression
Part of the abstract was presented at the 2016 AACR
Annual Conference, New Orleans, LA, USA, April 2016.
Author contributions: Ying Wang had full access to all the data in the
study and takes responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: Wang, Jacobs, Gapstur, Patel.
Acquisition of data: All authors.
Analysis and interpretation of data: All authors.
Drafting of the manuscript: Wang.
Critical revision of the manuscript for important intellectual content: Jacobs,
Gapstur, Gansler, McCullough, Stevens, Patel.
E U R O P E A N U R O L O G Y 7 2 ( 2 0 1 7 ) 9 3 1 – 9 3 9 937