Table 2.
Stratified analyses of pooled relative risks (95% CI) for cardiovascular and stroke outcomes (number of pooled studies in parentheses after each effect estimate)
| Cardiovascular disease mortality (n=21 studies, 1<thin>184<thin>956 subjects) | Coronary heart disease | Stroke | ||||
|---|---|---|---|---|---|---|
| Incident (n=29 studies, 549<thin>504 subjects) | Mortality (n=31 studies, 1<thin>925<thin>106 subjects) | Incident (n=17 studies, 458<thin>811 subjects) | Mortality (n=10 studies, 723<thin>571 subjects) | |||
| Active drinkers v non-drinkers: | ||||||
| Least adjusted models | 0.84 (0.75 to 0.95) (11) | 0.73 (0.65 to 0.82) (14) | 0.80 (0.70 to 0.91) (10) | 1.01 (0.88 to 1.16) (10) | 1.13 (0.96 to 1.32) (3) | |
| Most adjusted models | 0.75 (0.70 to 0.80) (21) | 0.71 (0.66 to 0.77) (29) | 0.75 (0.68 to 0.81) (31) | 0.98 (0.91 to 1.06) (17) | 1.06 (0.91 to 1.23) (10) | |
| Active drinkers v lifetime abstainers | 0.82 (0.78 to 0.86) (9) | 0.73 (0.61 to 0.88) (9) | 0.75 (0.66 to 0.85) (7) | 0.93 (0.85 to 1.02) (7) | 1.29 (1.09 to 1.53) (3) | |
| Former drinkers v non-drinkers | 1.48 (1.23 to1.79) (6) | 1.10 (0.91 to 1.33) (8) | 1.31 (1.02 to 1.68) (6) | 0.87 (0.72 to 1.07) (4) | Not reported (2) | |
| Alcohol intake (g/day) v none: | ||||||
| <2.5 | 0.71 (0.57 to 0.89) (7) | 0.96 (0.86 to 1.06) (6) | 0.92 (0.80 to 1.06) (6) | 0.81 (0.74 to 0.89) (3) | 1.00 (0.75 to 1.34) (3) | |
| 2.5–14.9 | 0.77 (0.71 to 0.83) (15) | 0.75 (0.65 to 0.88) (9) | 0.79 (0.73 to 0.86) (18) | 0.80 (0.74 to 0.87) (3) | 0.86 (0.75 to 0.99) (6) | |
| 15–29.9 | 0.75 (0.70 to 0.80) (13) | 0.66 (0.59 to 0.75) (15) | 0.79 (0.71 to 0.88) (15) | 0.92 (0.82 to 1.04) (5) | 1.15 (0.86 to 1.54) (6) | |
| 30–60 | 0.85 (0.73 to 0.98) (10) | 0.67 (0.56 to 0.79) (9) | 0.77 (0.72 to 0.83) (12) | 1.15 (0.98 to 1.35) (4) | 1.10 (0.85 to 1.45) (5) | |
| >60 | 0.99 (0.84 to 1.17) (6) | 0.76 (0.52 to 1.09) (9) | 0.75 (0.63 to 0.89) (9) | 1.62 (1.32 to 1.98) (4) | 1.44 (0.99 to 2.10) (3) | |
| Sex: | ||||||
| Men | 0.80 (0.73 to 0.87) (13) | 0.71 (0.66 to 0.77) (25) | 0.77 (0.72 to 0.82) (21) | 1.02 (0.92 to 1.13) (11) | 1.07 (0.89 to 1.28) (9) | |
| Women | 0.69 (0.60 to 0.78) (9) | 0.71 (0.66 to 0.77) (11) | 0.78 (0.64 to 0.94) (10) | 0.87 (0.75 to 1.01) (4) | 0.81 (0.67 to 0.98) (3) | |
| Adjustment for confounding factors*: | ||||||
| Weak | 0.74 (0.67 to 0.82) (10) | 0.69 (0.62 to 0.76) (11) | 0.72 (0.63 to 0.83) (15) | 0.99 (0.86 to 1.13) (7) | 1.30 (1.11 to 1.52) (5) | |
| Strong | 0.76 (0.70 to 0.83) (11) | 0.72 (0.65 to 0.79) (18) | 0.80 (0.75 to 0.86) (16) | 0.99 (0.89 to 1.09) (10) | 0.96 (0.81 to 1.14) (5) | |
| Median follow-up time†: | ||||||
| Short | 0.76 (0.71 to 0.83) (8) | 0.71 (0.65 to 0.79) (14) | 0.75 (0.67 to 0.85) (12) | 0.98 (0.90 to 1.07) (9) | 1.01 (0.82 to 1.24) (5) | |
| Long | 0.75 (0.67 to 0.84) (13) | 0.72 (0.64 to 0.80) (15) | 0.75 (0.67 to 0.84) (19) | 1.00 (0.88 to 1.13) (8) | 1.18 (1.02 to 1.37) (5) | |
*Adjustment for confounding factors was dichotomised as weak (<median value) or strong (≥median value). Cut points: ≥5 for coronary heart disease and stroke mortality, ≥6 for cardiovascular disease mortality and incident coronary heart disease, ≥7 for incident stroke.
†Total follow-up time was dichotomised as short (<median value) or long (≥median value). Cut points: ≥9 for incident coronary heart disease, ≥10 for cardiovascular disease mortality, ≥12 for coronary heart disease mortality and incident stroke, ≥14 for stroke mortality.