933 1030
2.3.
Outcomes
All deaths through December 31, 2012 were ascertained through linkage
with the National Death Index
[22] .PCSM was defined as deaths with
prostate cancer coded as the singular underlying cause of death per
International Classification of Diseases codes (ninth revision: 185; 10th
revision: C61). Cardiovascular disease (CVD) mortality and all-cause
mortality were examined as secondary outcomes.
2.4.
Statistical analysis
Cox proportional hazards regression models were used to calculate
cause-specific hazard ratios (HRs) and 95% CIs for PCSM, CVD mortality,
and all-cause mortality across categories of recreational PA. All deaths
other than the outcome were censored at death date. Follow-up time in
prediagnosis analyses started from diagnosis date and ended on death
date or December 31, 2012, whichever came first. Follow-up time in
postdiagnosis analyses started 4 yr after completion of the postdiagnosis
questionnaire.
All multivariable-adjusted models were stratified by single year of
age at diagnosis, and controlled for calendar year of diagnosis, race,
tumor extent, Gleason score, and nodal involvement. Prediagnosis
models further controlled for education, family history of prostate
cancer, and the following lifestyle factors as measured on the
prediagnosis survey: body mass index (BMI), smoking status, CVD
history, current multivitamin supplement use, red and processed meat
intake, fish intake, sitting time, and any history of PSA testing not leading
to prostate cancer diagnosis, categorized as shown in
Table 1 .Table 1 – Selected characteristics by prediagnosis recreational physical activity among men diagnosed with nonmetastatic prostate cancer in
the CPS-II Nutrition Cohort
aPrediagnosis recreational physical activity (MET-h/wk)
<
3.5
(
n
= 865)
N
(%)
3.5–8.75
(
n
= 2312)
N
(%)
8.75–
<
17.5
(
n
= 1614)
N
(%)
17.5
(
n
= 2537)
N
(%)
Median MET value
0
3.5
14
25.7
Age at diagnosis (yr)
<
65
128 (15.4)
326 (12.8)
205 (13.2)
261 (10.9)
65–
<
70
202 (24.3)
662 (27.5)
407 (25.6)
703 (28)
70–
<
75
226 (26)
724 (31.5)
512 (31.6)
823 (32.4)
75–
<
80
192 (22)
417 (19.3)
357 (21.7)
552 (21.1)
I
80
117 (12.3)
183 (8.9)
133 (8)
198 (7.5)
Diagnosis year
b1992–1998
384 (44.4)
1207 (52.2)
691 (42.8)
1045 (41.2)
1999–2005
317 (36.6)
830 (35.9)
690 (42.8)
1128 (44.5)
2006–2011
164 (19)
275 (11.9)
233 (14.4)
364 (14.3)
Race
White
839 (97)
2242 (96.9)
1567 (97.1)
2491 (98.2)
Black
18 (2.1)
41 (1.8)
22 (1.3)
28 (1.1)
Other/unknown
8 (0.9)
29 (1.3)
25 (1.6)
18 (0.7)
Tumor extent
T1–T2
791 (91.3)
2143 (93.1)
1501 (92.8)
2349 (92.4)
T3–T4
74 (8.7)
169 (6.9)
113 (7.2)
188 (7.6)
Nodal involvement
No
855 (98.8)
2281 (98.7)
1598 (99)
2515 (99.1)
Yes
10 (1.2)
31 (1.3)
16 (1)
22 (0.9)
Gleason score
2–6
400 (46.4)
1186 (51)
823 (51.1)
1313 (51.8)
7
218 (24.4)
529 (23.5)
418 (25.6)
628 (24.4)
8–10
136 (15.7)
274 (12)
179 (11)
319 (12.6)
5–7, 7
[7_TD$DIFF]
–10 or unknown
111 (13.4)
323 (13.4)
194 (12.2)
277 (11.2)
Education
<
High school
122 (14.1)
167 (7.1)
87 (5.4)
127 (5.1)
High school graduate
208 (24.6)
414 (18)
250 (15.5)
353 (13.9)
Some college
232 (26.9)
593 (25.6)
380 (23.6)
589 (23.3)
College graduate/graduate school
303 (34.4)
1138 (49.3)
897 (55.5)
1468 (57.7)
Family history of prostate cancer
No
818 (94.3)
2191 (94.8)
1536 (95.1)
2426 (95.6)
Yes
47 (5.7)
121 (5.2)
78 (4.9)
111 (4.4)
Any history of prediagnosis PSA testing not leading to prostate cancer diagnosis
c[6_TD$DIFF]
No
171 (20.4)
431 (16.9)
241 (15.6)
329 (13.9)
Yes
618 (70.5)
1674 (74.5)
1258 (77.1)
2027 (78.7)
Unknown
76 (9.1)
207 (8.6)
115 (7.3)
181 (7.4)
Prediagnosis history of cardiovascular disease
No
710 (82.6)
1895 (80.9)
1328 (82.6)
2062 (81.7)
Yes
155 (17.4)
417 (19.1)
286 (17.4)
475 (18.3)
Prediagnostic body mass index (kg/m
2
)
<
18.5
8 (0.9)
7 (0.3)
4 (0.3)
12 (0.5)
18.5–
<
25
239 (27.5)
738 (31.9)
575 (35.6)
1049 (41.6)
25–
<
30
416 (48.3)
1195 (51.4)
813 (50.3)
1199 (47.1)
30
192 (22.1)
348 (15.3)
208 (12.9)
254 (9.9)
Unknown
10 (1.2)
24 (1)
14 (0.9)
23 (0.9)
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
933