Table 1: Effect of Zinc Lozenges on Common Cold Episodes of Natural Origin

Study [Ref.] No. of Participants Zn Dose Per Day (mg)a Average Duration of Colds (Days)b Zn/Placebo The Effect of Zn Pb (1-Tail) –2×ln(P)
Eby et al.1984 [7] 65 207 3.9/10.8c -64%c 0.0005c 15.2
Smith et al.1989 [19] 110 207 5.9/6.3d
5.5/6.9d
-22%d 0.5d
0.01d
1.4
Godfrey et al. 1992 [20] 73 192 4.9/6.1 -21% 0.024 7.4
Prasad et al. 2008 [21] 50 92 4.0/7.1 -44% 6×10–13 56.3
Petrus et al. 1998 [22] 101 89 3.8/5.1 -26% 0.0033 11.4
Turner at al. A 2000 [23] 139 80 6.0/5.5 0.5 1.4
Mossad et al.1996 [24] 99 80 4.4/7.6 -42% 0.0005 15.2
Prasad et al. 2000 [25] 48 80 4.5/8.1 -44% 2×10–9 40.0
Turner et al. B 2000 [23] 139 69 5.5/5.5 0.5 1.4
Douglas et al. 1987 [26] 58e 64 12.1/7.7 0.96 0.1
Macknin et al. 1998 [27] 247 45 9.0/9.0 0.5 1.4
Weismann et al. 1990 [28] 130 45 7/6 0.5 1.4
Turner et al. C 2000 [23] 143 30 6.0/5.5 0.5 1.4

a Calculation of the daily dose of zinc: see Supplementary Material 2.
b The outcome is the mean or median of common cold duration, except when otherwise stated. The P-values were recalculated when appropriate data was reported in the paper.
c Eby et al. [7] did not report the mean or median duration, but estimated the time at which half of the participants were cured from an exponential fit of the results. The P-value at this table was calculated by using the Fisher exact test for the number of participants reporting no symptoms after the 7-day trial: 32/37(86%) and 13/28(46%) in the zinc and placebo groups, respectively.
d Smith et al. [19] reported that “subjects taking zinc gluconate had lower severity scores than those in the corresponding placebo group on days 4 to 7 of treatment. This difference is statistically significant (P = 0.02).” From Smith et al.’s fig. 2, the days needed for 80% reduction in the severity score, which occurred in the 4 to 7 day time range, was measured thereby transforming the effect to the time scale for this table. The upper line (5.9/6.3) gives the interpolated time point when 40% of participants had become asymptomatic from Smith et al.'s fig. 1; the placebo participants were not followed until half of them had become asymptomatic. To be conservative, P = 0.5 was used for calculating the –2×log(P), and not the small P-value corresponding to the difference in severity scores on days 4 to 7.
e The number of treatment courses was 63; some of the 58 participants had more than one cold episode.