Skip to main content
. Author manuscript; available in PMC: 2014 Sep 22.
Published in final edited form as: Diabetes Care. 2014 Mar 21;37(6):1712–1720. doi: 10.2337/dc13-1731

Table 2. Cox Proportional Hazard Ratios showing the association between health behaviour change and rate of CVD events over 5-years follow-up among newly diagnosed type 2 diabetes patients from Eastern England (ADDITION-Cambridge) 2002 to 2009.

Change Category* Health
Score
Mean (SD) n D Y Model 1 Model 2 Model 3

Rate Ratio(95%CI) p Rate Ratio(95%CI) p Rate Ratio(95%CI) p
Change in Physical Activity (Net MET hrs · day−1)
Decreased 0 − 4.2 (4.4) 361 23 1.87 1 0.37 1 0.037 1 0.037
Increased 1 + 4.2 (5.1) 380 23 1.93 0.76 (0.42 to 1.38) 0.53 (0.29 to 0.96) 0.49 (0.26 to 0.96)
Change in Alcohol intake (Units · week−1)
Consistent drinkers/Increased 0 + 2.9 (5.7) 262 25 1.34 1 0.02 1 0.008 1 0.009
Abstainers/Reduced 1 − 3.0 (5.6) 458 20 2.35 0.50 (0.27 to 0. 91) 0.40 (0.21 to 0.78) 0.38 (0.19 to 0.79)
Change in Diet Daily total calorific intake (Kcal · day−1)
Increased 0 +351 (368) 243 17 1.23 1 0.11
Decreased 1 −554 (474) 489 29 2.52 0.62 (0.35 to 1.11)
Daily total fat intake (% Energy from fat . day−1)
Increased 0 +5.1 (3.9) 268 17 1.37 1 0.84
Decreased 1 −5.4 (4.3) 464 29 2.38 1.08 (0.52 to 2.23)
Daily fibre intake (g · day−1)
Decreased 0 −4.3 (4.6) 307 17 1.57 1 0.47
Increased 1 +5.5 (5.2) 425 29 2.18 1.30 (0.64 to 2.66)
Plasma Vitamin C (μmol · L−1)
Decreased 0 −16.2 (13.9) 295 18 1.51 1 0.32 1 0.18 1 0.14
Increased 1 +17.7(16.6) 340 20 1.69 0.75 (0.44 to 1.31) 0.67 (0.38 to 1.20) 0.63 (0.34 to 1.17)
Smoking Status
Smoker NA NA 151 10 0.76 1 0.71
Never/former NA NA 703 43 3.59 0.88 (0.47 to 1.68)
BMI (kg/m2)
Increased - +0.90 (0.8) 183 11 0.93 1 0.82 - -
Decreased - −1.89 (1.7) 541 34 2.78 1.06 (0.61 to 1.84)
*

Change denotes an increase or decrease in health behaviour from baseline to one year of follow-up. All models adjusted for age, sex, baseline value of relevant ‘change’ variable and study group. Multivariable Model 1 presents the independent effects of each ‘change in health behaviour’. Model 2, as for Model 1 and additionally adjusted for social class, occupation and mutually adjusted for change in physical activity, alcohol intake, plasma vitamin C levels. Model 3, as for Model 2 with the addition of self-reported prescription of cardio-protective medication at 1 year. D: primary composite CVD outcome; Y: # of person years at risk (per 1000). Clustering by GP practise taken into account in all models. † at baseline (n). - omitted from multivariable analysis as may be on causal pathway. CVD, cardiovascular disease.