Table 1.

Effect of vitamin C on the proportion of participants with illness in the large Anderson (5) trial.

Outcome	Intervention		Effect of vitamin C (95% CI)	P (2-tail)
	Vitamin C	Placebo
No days confined indoors	232/407 (57%)	195/411 (47%)	9.6 (2.7, 17) pp	0.006
No days off work	275/407 (67%	243/411 (59%)	8.4 (1.8, 15) pp	0.012
No days with nose or throat symptoms	131/407 (32%)	101/411 (25%)	7.6 (1.4, 14) pp	0.016
Free of any illness during the trial	105/407 (26%)	76/411 (18%)	7.3 (1.6, 13) pp	0.012

Participants were administered 1 g/day vitamin C regularly and 3 g/day extra for three days when they were sick. The 9.6 percentage point difference in the proportion of participants who had no days confined indoors indicates that 9.6% of participants benefited from vitamin C on the basis of this outcome in the particular context of the Anderson (5) trial. Anderson restricted to persons who normally experienced at least one cold in the period January to March. The data for “days confined indoors” is from p. 506 and Table VI, the data for “days off work” is from p. 506, the data for the number of subjects who did not experience any nose or throat symptoms are from p. 508, and the data for any illness is from p. 505 of (5). An earlier version of this table was published as Table 20 in ref. (9). The calculations were done with the prop.test program of the R package.

pp, percentage point difference in the outcome between the two groups.