Vitamin C does not prevent colds. Zinc may shorten one, with more side effects.

The most recent Cochrane review of zinc opens with a sentence that answers the question before the evidence starts: "There are no established interventions to prevent colds or shorten their duration" [s2]. Everything below is a description of how close the best-studied candidates come to that bar, and the answer is that vitamin C does not prevent colds in ordinary circumstances, zinc may shorten one that has already started but on low-certainty evidence and with more side effects, and antibiotics do nothing at all.

Vitamin C: prevention no, duration slightly

The Cochrane review of vitamin C restricted itself to placebo-controlled trials using at least 0.2 g per day [s1]. Twenty-nine trial comparisons involving 11,306 participants contributed to the analysis of whether regular supplementation prevents colds [s1].

In the general community, covering 10,708 participants, it did not: the pooled risk ratio was 0.97 (95% CI 0.94 to 1.00) [s1]. There is one striking exception. Five trials involving 598 marathon runners, skiers and soldiers on subarctic exercises produced a pooled risk ratio of 0.48 (95% CI 0.35 to 0.64) — roughly halving the incidence of colds in people undergoing brief periods of severe physical stress [s1].

Duration is where the more commonly repeated claim comes from. Across 31 comparisons covering 9,745 cold episodes, regular vitamin C shortened colds by 8% in adults (95% CI 3% to 12%) and by 14% in children (95% CI 7% to 21%), with 1 to 2 g per day in children shortening colds by 18% [s1]. Severity was also reduced by regular supplementation [s1].

But the version of the claim most people act on — taking vitamin C once symptoms appear — was tested separately across seven comparisons and 3,249 episodes, and no consistent effect on duration or severity was seen [s1]. The reviewers' conclusion is that routine supplementation is not justified by the incidence data, while the duration finding from regular supplementation was not replicated in the small number of therapeutic trials that exist [s1].

An 8% reduction, for reference, is a few hours off a week-long illness.

Zinc: the effect is larger and the certainty is lower

The 2024 Cochrane review of zinc included 34 studies — 15 on prevention, 19 on treatment — involving 8,526 participants, 22 in adults and 12 in children [s2]. Most tested lozenges, with zinc gluconate the most common form, given at doses between 45 and 276 mg per day for between 4.5 and 21 days [s2]. Most studies were at unclear or high risk of bias in at least one domain [s2].

For prevention, zinc did little. There may be little or no reduction in the risk of developing a cold (risk ratio 0.93, 95% CI 0.85 to 1.01; 9 studies, 1,449 participants; low certainty), and when colds occurred there was probably little or no difference in duration (mean difference 0.63 days shorter, 95% CI 1.29 days shorter to 0.04 days longer; 3 studies; moderate certainty) [s2].

For treatment of a cold already underway, the pooled estimate is much larger: a mean reduction in duration of 2.37 days (95% CI 0.53 to 4.21 days shorter) across 8 studies and 972 participants [s2]. That estimate carries two heavy caveats. Its certainty is graded low, and the heterogeneity statistic is I² = 97%, meaning the individual trials disagree with each other almost completely [s2]. It is also uncertain whether zinc reduces the risk of still having a cold at the end of follow-up (risk ratio 0.52, 95% CI 0.21 to 1.27; very low certainty), and there may be little or no difference in global symptom severity [s2].

The harm side is graded more confidently than the benefit. There is probably an increase in non-serious adverse events when zinc is used to treat a cold (risk ratio 1.34, 95% CI 1.15 to 1.55; 16 studies, 2,084 participants; moderate certainty) [s2]. The events catalogued include unpleasant taste, loss of smell, vomiting, stomach cramps and diarrhoea [s2]. No treatment study reported information on serious adverse events at all [s2].

The reviewers close by noting the wide variation in interventions and outcomes across studies and incomplete reporting in several domains, which they say should be considered when drawing conclusions about efficacy [s2].

Antibiotics: the settled negative

For completeness, the intervention most often requested is the one with the clearest answer. Pooling six trials with 1,047 participants, people given antibiotics for a common cold did no better than those given placebo on lack of cure or persistence of symptoms (risk ratio 0.95, 95% CI 0.59 to 1.51), while the risk of adverse effects was 1.8 times higher (95% CI 1.01 to 3.21) [s3]. In adults specifically the risk of adverse effects was 2.62 times higher (95% CI 1.32 to 5.18) [s3].

What the evidence base looks like as a whole

Three features recur across these reviews and are worth more attention than any single effect size.

The trials are heterogeneous to the point of near-incoherence — I² of 97% for the zinc treatment result, and interventions ranging from lozenges to nasal sprays to syrups at doses spanning a six-fold range [s2]. The certainty grades attached to the positive findings are low or very low, while the grade attached to the increase in side effects is moderate [s2]. And the largest apparent benefit in the vitamin C literature — halving cold incidence — comes from a narrow population of people under extreme physical stress, not from ordinary life [s1].

None of this is advice about whether to take anything. Doses used in trials are not recommendations, zinc in particular has known adverse effects at the doses studied, and anyone weighing a supplement alongside other medicines or conditions should discuss it with a clinician or pharmacist.