Table 2

Number of events and hazard ratios for coronary heart disease and stroke by levels of baseline alcohol consumption (g/day)

CharacteristicNon-fatalFatal
EventsHazard ratio (95% CI)EventsHazard ratio (95% CI)
Coronary heart disease
Non-drinkers15921.15 (1.03 to 1.28)3321.25 (1.01 to 1.53)
0.1-4.927971 (ref)4971 (ref)
5.0-14.922070.82 (0.75 to 0.90)4180.83 (0.70 to 0.98)
15.0-29.913240.78 (0.70 to 0.87)1980.65 (0.53 to 0.81)
30.0-59.910270.73 (0.65 to 0.83)1740.82 (0.65 to 1.03)
≥60.03600.68 (0.57 to 0.81)800.98 (0.70 to 1.37)
P value*<0.0010.002
12 g/day increaseLinear0.94 (0.92 to 0.96)Linear0.92 (0.85 to 0.99)
Quadratic1.01 (1.00 to 1.02)
P value for trend<0.001P value0.003
Stroke
Non-drinkers9241.26 (1.12 to 1.43)1871.41 (1.12 to 1.79)
0.1-4.915731 (ref)2141 (ref)
5.0-14.915081.03 (0.93 to 1.14)1671.04 (0.83 to 1.31)
15.0-29.98721.08 (0.96 to 1.22)881.07 (0.81 to 1.42)
30.0-59.97041.10 (0.96 to 1.26)611.20 (0.87 to 1.67)
≥60.02741.31 (1.07 to 1.60)161.14 (0.65 to 2.01)
P value*0.1090.863
12 g/day increaseLinear1.04 (1.02 to 1.07)Linear1.05 (0.98 to 1.13)
P value for trend0.0020.136

Models were stratified by centre and sex, and systematic adjustment was undertaken for age at recruitment, body mass index, height, physical activity, smoking status, and history of hypertension.

  • * P value for the Wald test statistics compared with a χ2 distribution with four degrees of freedom, not including the category of non-drinkers (<0.1 g/day).

  • P value for baseline alcohol consumption modelled as a continuous variable, with inclusion in the model of an indicator variable expressing alcohol consumption at baseline.

  • P value for inclusion of a quadratic term for baseline alcohol consumption.