Table 1.

Characteristics of the trials included in the meta-analyses

Study (location)	Age of trial participants in years	Duration of follow-up	Infections assessed	No of subjects in placebo group	No of subjects in treatment group	Jadad score	Comments
Chavance et al 1993 (France)21	60+	4 months	Respiratory, nose, throat, ear, skin, mouth, urinary, and gynaecological infections	108 (of which 7 were lost to follow-up, leaving an effective sample size of 101)	110 (of which 7 were lost to follow-up, leaving an effective sample size of 103)	3	Double blind
Girodon et al 1999 (France)22	65+	2 years (maximum)	Respiratory and symptomatic urogenital infections	182 (of which 4 withdrew and 51 died during the study, but all were included in analysis)	181 (of which 3 withdrew and 55 died during the study, but all were included in analysis)	4	A double blind, 2×2 factorial trial (arms multivitamin only and minerals only excluded from meta-analyses) Population was from nursing homes
Chandra 1992 (United States)23	65+	12 months	Any infection (diagnosis was based on clinical features and laboratory tests included blood count, radiography of the chest and sinuses, bacterial and fungal cultures of sputum, urine and blood, C reactive protein, and erythrocyte sedimentation rate)	48	48	4	Double blind
Barringer et al 2003 (United States)24	65+	12 months	Upper respiratory tract infection, lower respiratory tract infection, influenza-like syndrome, gastrointestinal infection, and urinary tract infection	17	18 65 (2 dropped out and were lost to follow-up leaving an effective sample size of 16)	4	People aged <65 years were excluded from the analysis. A large proportion of subjects had diabetes. Double blind
Graat et al 2002, (Netherlands)25	60+	15 months (maximum)	Acute respiratory tract infections	230	240	3	A double blind, 2×2 factorial trial. Arms of vitamin E only and multivitamin+mineral + vitamin E have been excluded from meta-analyses
Chandra 2002 (location not specified)26	50-65	12 months	Common adult infection illnesses diagnosed by fever, cough, elevated erythrocyte sedimentation rate and C reactive protein, x ray of the sinuses and chest, blood culture, sputum culture, and urine culture	22 (three lost to follow-up, leaving an effective sample size of 19)	22	4	Double blind
Jain 2002 (India)27	51-78	12 months	Respiratory infections	18	18	2	Blinding unclear
Girodon et al 1997 (France)28	65+	2 years	Respiratory and symptomatic urogenital infections	20 (of which 7 dropped out before end but were included in analysis)	21 (of which 7 dropped out before end but were included in analysis)	4	A double blind, 2×2 factorial trial (arms multivitamin only and minerals only excluded from meta-analysis) Population was from nursing homes