Table 1.

Characteristics and results of the cohort studies stratified by source of activity (total, leisure time, occupational)

Author (reference)Cohort name and countryBaselineAge (y)Follow-up (y)Cohort sizeNo. casesActivity measure—life periodContrastRisk estimate (95% CI)*; trendQuality score (% of max)
Total activity
    Colbert et al. (13)Breast Cancer Detection Demonstration Project (United States)1987-198961 ± 88.223,369253 postmenoMET—recentHighest activity quintile vs lowestHR, 0.8 (0.5-1.1)75
Residual confounding: no
    Friberg et al. (6)Swedish Mammography Cohort199750-837.233,723199 postmenoMET—recentHighest activity quintile vs lowestRR, 0.79 (0.53-1.17); trend: no (P = 0.27)87
Residual confounding: partly
Leisure time activity
    Terry et al. (12)Swedish Twin Registry (Sweden)196742-8120.411,659133 pre- and post-menoSubjective—recentHard physical exercise vs none (4 cats)HR, 0.1 (0.04-0.6); trend: yes (P < 0.01)61
Residual confounding: yes
    Furberg and Thune (14)Norwegian Women Cohort (Norway)1974-198145 ± 715.724,460130 pre- and post-menoSubjective—recentGrade 3 + 4 (sports) vs grade 1 (sedentary)HR, 0.79 (0.43-1.45); trend: no (P = 0.39)75
Residual confounding: yes
    Schouten et al. (4)Netherlands Cohort Study on Diet and Cancer (Netherlands)198655-699.362,573226 postmenoDuration (h/wk)—recentTotal nonoccupational activity; ≥90 min/d vs <30 min/d (4 cats)HR, 0.54 (0.34-0.85); trend: yes (P = 0.002)68
Residual confounding: partly
    Friberg et al. (6)Swedish Mammography Cohort199750-837.233,723199 postmenoHours/d—recentHigh (≥20 min/d) vs low (<20 min/d)RR, 0.99 (0.73-1.32); trend: n.a.87
Residual confounding: partly
Occupational activity
    Pukkala et al. (15)Teachers cohort (Finland)1958-1973, 1984-199120-75+2410,11849 pre- and post-menoSubjective—lifetimePhysical exercise teachers vs language teachersSIRPE/SIRL: 0.85; trend: n.a.68
Residual confounding: partly
    Moradi et al. (16)Swedish Cancer Environment Registry (Sweden)1960-197016-9519989,2705,287 post- and post-menoSubjective—1970Sedentary vs high/very high (4 cats)RR, 1.32 (1.17-1.50); trend: yes (<0.001)71
Residual confounding: yes
    Furberg and Thune (14)Norwegian Women Cohort (Norway)1974-198145 ± 715.724,460130 pre- and post-menoSubjective—recentGrade 3 + 4 (manual) vs grade 1 (sedentary)HR, 0.61 (0.35-1.05); trend: no (P = 0.09)75
Residual confounding: yes
    Friberg et al. (6)Swedish Mammography Cohort199750-837.233,723199 postmenoMET—recentHigh (standing/manual) vs low (sedentary)RR, 1.01 (0.75-1.37); trend: n.a.87
Residual confounding: partly
  • * Printed in italic if contrast is low versus high activity levels.

  • Trend: subjective assessment (P value for trend test if available).