Skip to main content

Table 3 Best-fitting models and results of the meta-analysis on alcohol consumption and risk of mortality and subsequent cardiovascular events

From: Association of alcohol consumption with morbidity and mortality in patients with cardiovascular disease: original data and meta-analysis of 48,423 men and women

Outcome and subgroup

No. of studies (curves)

No. of patients

Maximal effect sizea

Reversion point, g/dayb

Powers for the Best-Fitting FP2

RR (95% CI)

g/day

dose_1

dose_2

All-cause mortality

 Overall

11 (11)

41,743

0.79 (0.73–0.85)

7

62

−0.5

1

 Male

6 (6)

19,897

0.82 (0.72–0.93)

9

39

0

0.5

 Female

3 (3)

6046

0.64 (0.36–1.14)

54

49

−2

3

 MI as primary event

9 (9)

29,554

0.82 (0.68–0.99)

2

7

−1

0.5

 Angina as primary event

2 (2)

8938

0.79 (0.63–0.99)

39

46

0.5

3

 Stroke as primary event

3 (3)

3618

0.71 (0.42–1.20)

12

NA

0

0.5

 Reference group including former drinkers

9 (9)

41,405

0.77 (0.69–0.85)

16

75

−0.5

2

 Reference group excluding former drinkers

4 (4)

17,526

0.85 (0.71–1.00)

3

3

−0.5

−0.5

 Post-event alcohol assessment

8 (8)

37,245

0.81 (0.74–0.88)

9

52

0

0.5

 Multiple alcohol measures

2 (2)

12,337

0.78 (0.59–1.03)

16

NA

−0.5

−0.5

Cardiovascular mortality

 Overall

9 (9)

24,770

0.73 (0.64–0.83)

8

50

0

0.5

 Male

5 (5)

14,536

0.72 (0.62–0.85)

9

32

0

0.5

 Female

2 (2)

4790

0.29 (0.09–1.01)

54

54

0

2

 MI as primary event

6 (6)

12,422

0.76 (0.64–0.91)

3

25

−2

3

 Angina as primary event

2 (2)

8934

0.72 (0.42–1.23)

56

NA

3

3

 Stroke as primary event

3 (3)

3617

0.63 (0.37–1.08)

26

NA

0

3

 Reference group including former drinkers

6 (6)

24,269

0.73 (0.58–0.93)

13

27

0

0.5

 Reference group excluding former drinkers

5 (5)

17,683

0.71 (0.55–0.90)

7

29

−0.5

0.5

 Post-event alcohol assessment

7 (7)

21,525

0.73 (0.60–0.90)

8

43

0

0

 Multiple alcohol measures

1 (1)

1818

0.58 (0.40–0.84)

17

33

−0.5

3

Cardiovascular events

 Overallc

4 (5)

28,621

0.50 (0.26–0.96)

6

15

−2

−2

 Male

3 (4)

13,598

0.56 (0.23–1.34)

8

NA

−2

−2

 Female

1 (1)

3775

0.67 (0.43–1.05)

54

49

−2

3

 MI as primary event

4 (5)

20,361

0.79 (0.66–0.94)

11

35

−2

3

 Angina as primary event

1 (1)

8747

0.69 (0.59–0.81)

35

n.a.

−2

1

 Stroke as primary event

1 (1)

1855

0.49 (0.26–0.92)

72

n.a.

−2

3

 Reference group including former drinkers

3 (3)

25,983

0.72 (0.53–0.97)

40

45

1

1

 Reference group excluding former drinkers

2 (3)

17,020

0.78 (0.46–1.31)

17

NA

3

3

 Multiple alcohol measures

1 (1)

353

0.32 (0.14–0.71)

38

n.a.

2

3

  1. FP2 second-degree fractional polynomial model, MI myocardial infarction
  2. aDefined as the lowest point of the dose-response curve within the range of dose reported by the studies
  3. bDefined as the dose of alcohol at which protection against the outcome is no longer statistically significant at the 95% confidence level; not applicable (NA) if non-significant association was found at any level of consumption; not available (n.a.) if the association remained significant within the range of dose reported by the studies
  4. cAll of the four studies measured post-event alcohol consumption and had a quality score ≥ 7
Back to article page

BMC Medicine

ISSN: 1741-7015

Contact us