Table 2.

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

Author (reference)CountryActivity assessmentAge (y)No. casesNo. controlsStudy base (cases/controls)Activity measure—life periodsContrastRisk estimate (95% CI)* trendQuality score (% of max)
Total activity
    Sturgeon et al. (17)United States1987-199020-74405 pre- and post-meno297Hospital based/population basedSubjective (global assessment)—recentVery inactive vs very active (5 cats)OR, 2.5 (0.7-8.7); trend: no (P = n.a.)60
Residual confounding: no
    Shu et al. (18)China1988-199018-74268 pre- and post-meno268Population based/population basedSubjective—age 50-59Very inactive vs very active (5 cats)OR, 2.1 (0.6-7.1); trend: no (P = n.a.)80
Residual confounding: no
    Levi et al. (19)Switzerland/Italy1988-199131-75274 pre- and post-meno572Hospital based/hospital basedSubjective—age 55Lowest vs highest (4 cats)OR, 8.6 (3.0-25.3); trend: no (P < 0.01)42
Residual confounding: yes
    Salazar-Martinez et al. (23)Mexico1995-1997?85 pre- and post-meno668Hospital based/hospital basedMET-h/wk—recent? leisure?≥38 MET-h/wk vs ≤29 (3 cats)OR, 0.47 (0.26-0.86); trend: no (P = 0.01)44
Residual confounding: yes
Leisure time activity
    Sturgeon et al. (17)United States1987-199020-74405 pre- and post-meno297Hospital based/population basedSubjective—recentLowest quintile vs highestOR, 1.2 (0.7-2.2); trend: no (P = n.a.)60
Residual confounding: no
    Shu et al. (18)China1988-199018-74268 pre- and post-meno268Population based/population basedMET—age 50-59Lowest quartile vs highest quartileOR, 1.0 (0.6-1.7); trend: no (P = n.a.)80
Residual confounding: no
    Levi et al. (19)Switzerland/Italy1988-199131-75274 pre- and post-meno572Hospital based/hospital basedSubjective—recentSporsts and leisure: lowest vs highest (4 cats)OR, 1.9 (0.9-4.0); trend: no (P < 0.01)42
Residual confounding: yes
    Hirose et al. (20)Japan1988-199318-80+145 pre- and post-meno26,751Hospital based/hospital basedFrequency—recentExercise for health: ≥3-4×/wk vs no activity (3 cats)OR, 0.63 (0.34-1.11); trend: no (P = n.a.)48
Residual confounding: yes
    Olson et al. (21)United States1986-199140-85232 pre- and post-meno631Hospital based/population basedDuration—recentVigorous activity: ≥100 h/y vs none (3 cats)OR, 0.67 (0.42-1.09); trend: yes (P = 0.06)47
Residual confounding: yes
    Goodman et al. (22)Hawaii (United States)1985-199318-84332 pre- and post-meno511Population based/population basedDuration (lifetime h)—lifetimeLifetime hours: >3,339 vs 0 h (4 quartiles)OR, 0.9 (CI incl 1); trend: no (P = 0.34)55
Residual confounding: yes
    Moradi et al. (24)Sweden1994-199550-74709 postmeno3,368Population based/population basedDuration—recentExercise/sports: 0 h/wk vs >2 h/wkOR, 1.3 (1.1-1.7); trend: yes (P = 0.01)70
Residual confounding: partly
    Littman et al. (25)United States1985-199145-74822 mostly postmeno?1,111Population based/population basedDuration—recentAny recreational activity: >6 h/wk vs none (4 cats)OR, 0.83 (0.59-1.15); trend: no (P = 0.31)58
Residual confounding: yes
    Matthews et al. (5)China1997-200130-69832 pre- and post-meno846Population based/population basedCalorie expenditure—lifetime (adult + adolescence)Highest quartile vs none (5 cats)OR, 0.60 (0.42-0.86); trend: yes (P < 0.01)74
Residual confounding: no
Occupational activity
    Sturgeon et al. (17)United States1987-199020-74405 pre- and post-meno297Hospital based/population basedFrequency—recentRarely/never vs dailyOR, 1.4 (1.0-2.1); trend: no60
Residual confounding: no
    Shu et al. (18)China1988-199018-74268 pre- and post-meno268Population based/population basedCaloric expenditure—lifetimeLowest quartile vs highest quartileOR, 1.0 (0.6-1.7); trend: no80
Residual confounding: no
    Levi et al. (19)Switzerland/Italy1988-199131-75274 pre- and post-meno572Hospital based/hospital basedSubjective—recentVery low vs high (4 cats)OR, 1.5 (1.0-2.2); trend: no (P < 0.05)42
Residual confounding: yes
    Zheng et al. (26)China1980-1984≥30452 pre- and post-meno?Population based/population basedJob category—recentHigh (>12 kJ/m) vs Low (<8 kJ/m)SIRhigh/SIRlow: 0.72; trend: yes55
Residual confounding: yes
    Dosemeci et al. (27)Turkey1979-1984all31 pre- and post-meno244Hospital based/hospital basedJob category—lifetimeSedentary (<8 kJ/m) vs active (>12 kJ/m)OR, 0.5 (0.0-9.3); trend: yes (P = 0.27)33
Residual confounding: yes
    Kalandidi et al. (28)Greece1992-1994all145 pre- and post-meno298Hospital based/hospital basedJob category—recentManual vs nonmanual (3 cats)OR, 0.41 (0.18-0.91); trend: yes45
Residual confounding: no
    Olson et al. (21)United States1986-199140-85232 pre- and post-meno631Hospital based/population basedCaloric expenditure—recentHighest vs lowest tertileOR, 1.19 (0.76-1.87); trend: no (P = 0.39)47
Residual confounding: yes
    Goodman et al. (22)Hawaii (United States)1985-199318-84332 pre- and post-meno511Population basedPopulation basedDuration (lifetime hours)—lifetimeLifetime hours: >20,089 h vs 0 h (4 quartiles)OR, 0.7 (CI incl 1); trend: no (P = 0.08)55
Residual confounding: yes
    Moradi et al. (24)Sweden1994-199550-74709 postmeno3,368Population based/population basedJob category—recentSedentary vs high/very highOR, 1.3 (0.9-1.9); trend: no (P = 0.05)70
Residual confounding: partly
    Matthews et al. (5)China1997-200130-69832 pre- and post-meno846Population based/population basedCalorie expenditure—lifetimeHighest quartile vs noneOR, 0.86 (0.63-1.18); trend: no (P = 0.70)74
Residual confounding: no
  • Abbreviation: n.a., not applicable; SIR, standard incidence ratio; PE, physical exercise teachers; L, language teachers; high: <12 kJ/m; low: <8 kJ/m.

  • * Printed in italic if contrast is low versus high activity levels.

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

  • Includes only postmenopausal women.