Skip to main content

Table 4 Sensitivity analyses for the association between leisure-time sunlight exposure and prostate cancer, PROtEuS, Montreal, Canada

From: Sunlight exposure during leisure activities and risk of prostate cancer in Montréal, Canada, 2005–2009

 

Cases

Controls

 

Sensitivity analysis

n

(%)

n

(%)

OR (95% CI)a

Excluding proxy respondents

1854

(100)

1889

(100)

 

CEvents b

     

  None: 0 event

55

(3.0)

46

(2.4)

1.45 (0.89-2.38)

  Q1: 1 to 3,210 events

490

(26.4)

456

(24.1)

1.13 (0.91-1.40)

  Q2: 3,211 to 6,642 events

431

(23.2)

463

(24.5)

0.94 (0.76-1.16)

  Q3: 6,643 to 13,679 events

464

(25.0)

460

(24.4)

1.01 (0.82-1.24)

  Q4: 13,680 to 94,886.67 events

414

(22.3)

464

(24.6)

1.00 (reference)

     

p-trend =0.52

CDuration c

     

  None: 0 hour

55

(3.0)

46

(2.4)

1.52 (0.94-2.50)

  Q1: 1 to 4,497 hours

466

(25.1)

457

(24.2)

1.14 (0.91-1.42)

  Q2: 4,498 to 9,711 hours

459

(24.8)

465

(24.6)

1.01 (0.82-1.25)

  Q3: 9,712 to 18,459 hours

480

(25.9)

457

(24.2)

1.18 (0.96-1.46)

  Q4: 18,460 to 103,484 hours

394

(21.3)

464

(24.6)

1.00 (reference)

     

p-trend =0.35

Excluding controls not screened for prostate cancer

1904

(100)

1491

(100)

 

CEvents b

     

  None: 0 event

55

(2.9)

34

(2.3)

1.39 (0.86-2.27)

  Q1: 1 to 3,210 events

506

(26.6)

336

(22.5)

1.12 (0.91-1.39)

  Q2: 3,211 to 6,642 events

438

(23.0)

373

(25.0)

0.88 (0.72-1.09)

  Q3: 6,643 to 13,679 events

474

(24.9)

374

(25.1)

1.03 (0.84-1.26)

  Q4: 13,680 to 94,886.67 events

431

(22.6)

374

(25.1)

1.00 (reference)

     

p-trend =0.56

CDuration c

     

  None: 0 hour

55

(2.9)

34

(2.3)

1.49 (0.92-2.45)

  Q1: 1 to 4,497 hours

483

(25.4)

339

(22.7)

1.18 (0.95-1.46)

  Q2: 4,498 to 9,711 hours

467

(24.5)

370

(24.8)

1.00 (0.81-1.23)

  Q3: 9,712 to 18,459 hours

489

(25.7)

365

(24.5)

1.19 (0.97-1.46)

  Q4: 18,460 to 103,484 hours

410

(21.5)

383

(25.7)

1.00 (reference)

     

p-trend =0.32

Excluding subjects without information for number of months per year of leisure activities

1611

(100)

1766

(100)

 

CEvents b

     

  None: 0 event

55

(3.4)

52

(2.9)

1.48 (0.92-2.41)

  Q1: 1 to 3,210 events

428

(26.6)

428

(24.2)

1.17 (0.93-1.47)

  Q2: 3,211 to 6,642 events

381

(23.6)

434

(24.6)

0.97 (0.78-1.21)

  Q3: 6,643 to 13,679 events

396

(24.6)

414

(23.4)

1.07 (0.86-1.32)

  Q4: 13,680 to 94,886.67 events

351

(21.8)

438

(24.8)

1.00 (reference)

     

p-trend =0.41

CDuration c

     

  None: 0 hour

55

(3.4)

52

(2.9)

1.55 (0.96-2.52)

  Q1: 1 to 4,497 hours

397

(24.6)

429

(24.3)

1.12 (0.89-1.42)

  Q2: 4,498 to 9,711 hours

407

(25.3)

428

(24.2)

1.09 (0.87-1.36)

  Q3: 9,712 to 18,459 hours

426

(26.4)

428

(24.2)

1.24 (0.99-1.54)

  Q4: 18,460 to 103,484 hours

326

(20.2)

429

(24.3)

1.00 (reference)

     

p-trend =0.30

Considering winter leisure activities considered as non-exposed

1904

(100)

1962

(100)

 

CEvents b

     

  None: 0 event

60

(3.2)

54

(2.8)

1.47 (0.93-2.35)

  Q1: 1 to 2,173 events

499

(26.2)

477

(24.3)

1.19 (0.96-1.48)

  Q2: 2,174 to 4,560 events

430

(22.6)

477

(24.3)

0.97 (0.79-1.20)

  Q3: 4,561 to 9,103 events

525

(27.6)

477

(24.3)

1.24 (1.01-1.52)

  Q4: 9,104 to 52,991.5 events

390

(20.5)

477

(24.3)

1.00 (reference)

     

p-trend =0.34

CDuration c

     

  None: 0 hour

60

(3.2)

54

(2.8)

1.38 (0.87-2.20)

  Q1: 1 to 2,987 hours

455

(23.9)

477

(24.3)

1.03 (0.83-1.28)

  Q2: 2,988 to 6,790 hours

474

(24.9)

477

(24.3)

1.01 (0.82-1.24)

  Q3: 6,791 to 12,479 hours

492

(25.8)

476

(24.3)

1.14 (0.93-1.40)

  Q4: 12,480 to 94,933 hours

423

(22.2)

478

(24.4)

1.00 (reference)

     

p-trend =0.46

Using self-reported recreational sunlight exposure

1904

(100)

1962

(100)

 

  None

105

(5.5)

117

(6.0)

0.98 (0.71-1.37)

  Sometimes

708

(37.2)

804

(41.0)

0.91 (0.79-1.06)

  Often

1091

(57.3)

1041

(53.1)

1.00 (reference)

  1. aOdds ratio (OR), 95% confidence interval (CI). Models adjusted for age, first-degree family history of prostate cancer, ancestry, timing of last prostate cancer screening, education, BMI, solar protection during leisure and overall physical activity level. Analyses excluding controls not screened for prostate cancer in the previous 2 years are not adjusted for prostate cancer screening. bCumulative number of events entailing sunlight exposure from age 18 years. cCumulative duration of leisure-time exposure to sunlight from age 18 years.