2. Characteristics of included trials with high risk of bias.

Trial	Design	Number participants	Mean age	Suppl (Y)	Beta‐carotene (mg)	Vitamin A (IU)	Vitamin C (mg)	Vitamin E (IU)	Selenium (µg)
Gillilan 1977	Crossover	52	57	0.5				1600
Mckeown‐Eyssen 1988	Parallel	185	58	2			400	400
Burns 1989	Parallel	19	81	0.1			200
Penn 1991	Parallel	30	84	0.077		8000	100	50
Chandra 1992	Parallel	96	74	1	16	1333	80	44	20
NIT1 1993	1/2 (2 x 2 x 2 x 2)	29584	NA	5.25	15	5000	120	33	50
de la Maza 1995	Parallel	74	50	1				500
Takamatsu 1995	Parallel	147	47	6				136
ter Riet 1995	2 x 2	88	NA	0.23			1000
Hogarth 1996	2 x 2	106	83	0.083		8000	500
ADCS 1 1997	2 x 2	341	73	2				2000
Girodon 1997	2 x 2	81	84	2	6		120	15	100
Bonelli 1998	Parallel	304	NA	5		6000	180	30	200
GISSI 1999	2 x 2	11324	59	3.5				330
de Waart 2001	Parallel	218	60	1.8				400
PPP 2001	2 x 2	4495	64	3.6				330
Stevic 2001	Parallel	28	57	1				1200	31.5
Sasazuki 2003	2 x 2	439	57	5	15		500
Takagi 2003	Parallel	93	63	5				600
ADCS 2 2005	Parallel	516	73	3				2000
SIT 2006	2 x 2 x 2	3411	NA	3.25	15		500	200	75
CTNS 2008	Parallel	1020	68	9		5000	60	30	25

Abbreviation: 
 NA, not available. 
 Blank cells indicate that the supplement was not part of the trial. 
 The years of follow‐up are the same as the years of supplementation.