Fig 1.
Flow chart of selection, randomization, and follow-up of participants approached and included in the analysis.
Table 1.
Baseline characteristics of study participants.
Fig 2.
The proportions of participants reporting a URI among those who completed the weekly surveys.
The ORs of reporting URI between the treatment arm and the placebo arm was obtained from a generalized linear mixed model adjusting for possible temporal trend, follow-up calendar year, and multiple responses per participant.
Table 2.
Odds ratios of reporting a symptom between the treatment and placebo groups for seven symptoms of interest.
Fig 3.
The weekly durations between two treatment groups by symptom.
Data for the placebo arm are in black while data for the treatment arm are in red. For each of the seven symptoms of interest, the p-values were obtained by comparing the mean differences of weekly durations between two treatment groups, using mixed-effect regression models with adjustments for multiple responses per participant, linear trend in time, and difference in calendar years.
Fig 4.
The weekly severity scores between two treatment groups by symptom.
Data for the placebo arm are in black while data for the treatment arm are in red. For each of the seven symptoms of interest, the p-values were obtained from comparing the mean differences of weekly severity scores between two treatment groups, using mixed-effect regression models with adjustments for multiple responses per participant, linear trend in time, and difference in calendar years. Note that individual severity score for fever had only two levels (0 or 1), while severity score for the other six symptoms had four levels (0, 1, 2, and 3).