Skip to main content

Table 3 Cox proportional hazards of incident CC according to BMI, sedentary behaviour and types of physical activity

From: Physical activity, obesity and sedentary behaviour and the risks of colon and rectal cancers in the 45 and up study

Variable Events No person-years Age-adjusted HR Multivariable HR(1)
BMI Kg/m2
 15- < 23.6 195 154,782 1.00 1.00
  ≥ 23.6- < 26.2 188 150,904 0.99 (0.81–1.21) 0.90 (0.73–1.12)
  ≥ 26.2- < 29.4 223 156,854 1.13 (0.93–1.37) 1.12 (0.91–1.38)
  ≥ 29.4-≤50 232 156,109 1.18 (0.98–1.43) 1.32 (1.08–1.63)
 Unknown 8 3273 1.94 (0.96–3.94) 2.21 (1.09–4.49)
P-value(a)     0.003
 P-trend(b)     0.0006
Sitting hours/day
 0- < 3 105 87,149 1.00 1.00
 3- < 5 228 167,892 1.13 (0.90–1.42) 0.98 (0.77–1.25)
 5- < 8 267 178,262 1.24 (0.99–1.56) 1.12 (0.88–1.42)
 8+ 165 143,987 0.95 (0.75–1.22) 1.02 (0.79–1.32)
 Unknown 81 44,633 1.51 (1.13–2.02) 0.92 (0.67–1.28)
P-value(a)     0.55
 P-trend(b)     0.88
Moderate activity minutes/week
 None 52 30,194 1.00 1.00
  > 0–149 175 137,275 0.73 (0.54–1.00) 1.08 (0.76–1.54)
 150 or more 583 433,966 0.78 (0.58–1.03) 1.25 (0.90–1.74)
 Unknown 36 20,487 1.01 (0.66–1.55) 1.09 (0.65–1.82)
P-value(a)     0.17
 P-trend(b)     0.76
Vigorous activity minutes/week
None 423 270,374 1.00 1.00
Any amount:(*) 220 248,354 0.56 (0.48–0.66) 0.78 (0.65–0.93)
  > 0–74 97 102,607 0.60 (0.48–0.75) 0.78 (0.61–0.98)
 75 or more 123 145,747 0.54 (0.44–0.66) 0.78 (0.63–0.97)
 Unknown 203 103,194 1.26 (1.07–1.49) 0.98 (0.81–1.18)
P-value(a)     0.02
 P-trend(b)     0.11
  1. (1)Multivariable models adjusted for: birth cohort, sex, education, BMI, sitting time, time spent on moderate and vigorous activity, smoking, alcohol, country of birth, guidelines of fruit and vegetables, weekly intake of processed food, red meat and fibre, aspirin, parental history of CRC and history of colorectal testing
  2. (a)P-value’ for each variable corresponds to a test of whether all HRs = 1
  3. (b)‘P-trend’ is for test of linear association in the log hazard scale and obtained by substituting the categorical versions of covariates in the Cox model with continuous or ordinal versions where appropriate
  4. ‘Unknown’ category was excluded from the estimation of P-values and P-trends.
  5. (*)The “any amount” category was derived by combining the “> 0–74” and “75 or more” and the HR for “any amount” was estimated by fitting a separate multivariate model