Table 2.

Hazard ratios for major cardiovascular outcomes in current drinkers, without and with adjustment for usual levels of systolic blood pressure, high-density-lipoprotein cholesterol, or body-mass index

	All stroke	Myocardial infarction	Coronary disease excluding myocardial infarction	Heart failure	Deaths from other types of cardiovascular disease
Subset of participants with measurement of systolic blood pressure
Cohorts/events	70/11 297	73/13 519	46/7789	39/2668	44/1019
Basic adjustment*	1·16 (1·11–1·22)	0·95 (0·91–0·99)	1·06 (1·00–1·12)	1·11 (1·04–1·18)	1·16 (1·06–1·27)
Plus adjustment for systolic blood pressure	1·10 (1·06–1·14)	0·91 (0·87–0·94)	1·03 (0·97–1·10)	1·08 (1·02–1·15)	1·14 (1·03–1·25)
Subset of participants with measurement of high-density-lipoprotein cholesterol
Cohorts/events	56/7982	61/9911	36/3608	29/1886	34/690
Basic adjustment*	1·16 (1·10–1·23)	0·93 (0·88–0·97)	1·07 (0·98–1·17)	1·09 (1·00–1·19)	1·22 (1·06–1·40)
Plus adjustment for high-density-lipoprotein cholesterol	1·17 (1·11–1·22)	1·00 (0·96–1·04)	1·13 (1·05–1·22)	1·14 (1·01–1·27)	1·22 (1·08–1·38)
Subset of participants with measurement of body-mass index
Cohorts/events	68/11 733	71/14 217	43/7761	36/2566	42/1035
Basic adjustment*	1·15 (1·10–1·19)	0·95 (0·91–0·98)	1·06 (1·02–1·12)	1·12 (1·04–1·20)	1·16 (1·06–1·27)
Plus adjustment for body-mass index	1·14 (1·10–1·18)	0·94 (0·91–0·97)	1·06 (1·01–1·12)	1·10 (1·03–1·16)	1·16 (1·06–1·27)

Data are hazard ratio (95% CI) per 100 g per week higher usual alcohol consumption, unless otherwise indicated. Analyses were restricted to individuals with basic adjustment variables plus the additional variable. Studies with fewer than five events were excluded from the analysis of each outcome.

^*Basic adjustment includes age, smoking, and history of diabetes, and stratification by sex and centre.