Substance + claimed effects

Vitamin C (ascorbic acid) is a water-soluble essential nutrient, taken as a supplement (tablets, lozenges, effervescent powder) at doses ranging from the dietary RDA of 75–90 mg/day up to mega-doses of 1–8 g/day for the express purpose of preventing or shortening the common cold NIH ODS 2021. Two distinct use patterns are evaluated separately in the literature, and they have different evidence bases: (1) regular prophylactic supplementation — taking it every day, all season, in the hope of catching fewer colds or having shorter ones when colds occur; and (2) acute therapeutic dosing at symptom onset — starting (or escalating) the dose the moment the first sore throat appears, in the hope of aborting or shortening the episode. The popularised claim that high-dose vitamin C broadly prevents colds traces to Pauling 1970; the question of whether that claim survived 60 years of trials is the substance of this entry. The entry covers prevention, symptom duration, symptom severity, the prophylactic-vs-acute contrast, dose-response, the special case of physically stressed populations, and the tolerable upper intake.

Evidence by addressing question

Mechanism

Science. Vitamin C is a cofactor for at least eight enzymes and a major aqueous-phase antioxidant. It is concentrated in leukocytes — neutrophil intracellular concentrations are roughly 50–100× plasma — and is consumed during the respiratory burst of phagocytosis, supporting chemotaxis, phagocyte function, and microbial killing Hemilä 2017. Plasma levels fall during acute infection, which is consistent with increased metabolic demand. The mechanistic case for an immune role is real and uncontroversial; the mechanistic case for a clinically meaningful cold effect at supplemental doses, in adults already at biochemical sufficiency, is weaker and inferential. Hemilä 2017 notes that animal evidence (148 studies in species that synthesise their own vitamin C) shows protective effects against bacterial, viral, and protozoal infections, raising the priors that there is some effect in humans — not that the effect is large.

Evidence

Science — prevention. The definitive synthesis is the Cochrane review by Hemilä & Chalker 2013: 29 placebo-controlled trials of regular vitamin C at ≥0.2 g/day, 11,306 participants in the general community. Pooled risk ratio for catching at least one cold: 0.97 (95% CI 0.94–1.00). Plain reading: regular supplementation does not reduce cold incidence in the ordinary adult population. This is a large-N, well-replicated null. The same review pooled 31 comparisons of cold duration (9,745 episodes) during regular use and found a small, consistent reduction: 8% in adults (95% CI 3–12%), 14% in children (95% CI 7–21%), with children at 1–2 g/day showing an 18% reduction.

Science — severity. The follow-up meta-analysis by Hemilä & Chalker 2023 (10 RCTs at ≥1 g/day, ~5,100 participants, 1,739–4,437 episodes per outcome) re-cut the data by severity and found a 15% reduction in overall cold severity (95% CI 9–21%). When trials separated severe from mild symptoms, the effect concentrated on the severe end — about a 26% reduction in severe-symptom days — with little effect on mild symptoms. Clinically this matters: it suggests vitamin C does not shave a few sniffles off a head cold but may compress the bedridden, fevered, indoors-confined fraction.

Science — therapeutic dosing at onset. The Cochrane review pooled 7 comparisons of vitamin C started after symptoms (3,249 episodes): no consistent benefit on duration Hemilä & Chalker 2013. Two trials (Anderson, in the Cochrane analysis) showed a dose-response within therapeutic use — 6 g/day shortened colds by 17% (twice the 3 g/day effect); 8 g/day on day 1 shortened colds by 19% (twice the 4 g/day effect) — suggesting the optimum, if there is one, lies above the doses ever formally tested Hemilä 2017. A 2018 meta-analysis that pooled "extra dose on top of routine" trials reported a half-day reduction, but was subsequently retracted for double-counted placebo arms; corrected estimates lost statistical significance.

Science — physical-stress subgroup. Five trials in marathon runners, skiers, and soldiers on subarctic exercises (598 participants total) found a pooled risk ratio of 0.48 (95% CI 0.35–0.64) — vitamin C halved cold incidence in this subgroup Hemilä & Chalker 2013. This is one of the most robust effects in the field; the contrast with the general-population null is the central interpretive puzzle.

Practice / clinical consensus. The Linus Pauling Institute — the institutional home of the pro-vitamin-C position — concludes that routine supplementation reduces incidence and duration in certain individuals, that benefit accrues only to prior regular use (not to dosing started after symptoms), and recommends ~≥1 g/day with a 2 g/day ceiling. The NIH Office of Dietary Supplements position is more conservative: routine supplementation is not generally justified to prevent colds, may shorten duration modestly, and the 2 g/day Tolerable Upper Intake Level applies. Mainstream clinical guidance (Mayo Clinic, USPSTF on multivitamins) does not recommend vitamin C for cold prevention.

Community / lay evidence. Vitamin C at the first sniffle is one of the most entrenched lay practices in the catalogue: emergency-room IV megadose marketing, effervescent "immune support" tablets, citrus folk remedies all reinforce it. Survey data and pharmacy sales spike during cold season. The perceived effect is widespread; the controlled effect, as above, is small and concentrated on regular use, not the at-onset pattern most lay users actually follow. The gap between lay belief and trial evidence is one of the largest in the supplement literature.

Protocol

For any benefit at all, regular use is what the trials support — 200 mg to 1 g/day daily, taken throughout cold season (or year-round). A reasonable upper bound is the FNB Tolerable Upper Intake Level of 2 g/day for adults NIH ODS 2021; above this, gastrointestinal side effects (osmotic diarrhea, cramping, nausea) rise sharply. Higher doses on the first day of symptoms (4–8 g) have a dose-response signal in two trials Hemilä 2017 but the at-onset benefit in pooled trials is null and starting from zero at symptom onset does not show benefit at all. The honest reading: routine 200–500 mg/day is harmless and may shorten colds by a small amount; megadosing at onset has weak signal and real GI risk.

Contraindications

Science. The principal supplementation hazards at high dose are osmotic diarrhea / GI upset (the basis for the 2 g/day UL) and, in susceptible men, an approximate doubling of kidney stone risk. Thomas et al. 2013, a prospective cohort of 23,355 Swedish men, found that supplemental ascorbic acid roughly doubled incident kidney stone risk (RR ~1.66 at ≤7 tablets/week; RR ~2.23 at >7/week); the mechanism is established — vitamin C is metabolised to oxalate, and ~2 g/day raises urinary oxalate by ~22%. The cohort was male, so the effect in women is less certain (smaller cohort signals exist). Independent risk groups: hereditary hemochromatosis (high-dose vitamin C enhances non-heme iron absorption, worsening iron overload), kidney disease (impaired oxalate clearance), and on a different timescale, possible interference with chemotherapy efficacy via antioxidant action (oncology guidance is to consult the treating clinician). Dietary vitamin C from fruit and vegetables is not implicated — the stone-formation signal is specific to supplemental high doses NIH ODS 2021.

Misconceptions

Four widely-believed-but-wrong claims, each anchored in the trial data:

• "Vitamin C prevents colds." False in the general adult population: the 11,306-participant Cochrane pool gives RR 0.97 (95% CI 0.94–1.00) — a null at the population level Hemilä & Chalker 2013. True only for athletes / soldiers under sustained acute physical stress (RR 0.48).
• "Take vitamin C when you feel a cold coming on." The intuitive pattern is the one with the weakest evidence. Pooled trials of dosing started after symptoms find no consistent duration reduction Hemilä & Chalker 2013. Whatever modest effect exists belongs to prior regular use, not a rescue dose.
• "More is better — megadose it." The dose-response signal in Anderson's trials suggests an optimum somewhere above 8 g/day was never tested, but the marginal benefit is small and the floor is firm: above 2 g/day GI side effects mount, and in male stone-formers high-dose supplementation roughly doubles kidney-stone incidence Thomas et al. 2013.
• "Studies have proven Pauling wrong." Overstated; the Karlowski 1975 trial — the most cited "vitamin C does nothing" result — had a documented blinding failure (placebo lactose tasted different from ascorbic acid), and reanalysis of the truly-blinded subgroup left a small genuine effect intact Karlowski et al. 1975. Pauling was wrong about the magnitude (and about cancer); he was not entirely wrong about a real, small biological effect.

Audience

Three reader groups behave differently and the entry should respect that. (1) The ordinary adult. Routine 200–500 mg/day is a near-zero-cost, near-zero-risk hedge; the expected benefit is modest. Megadosing at onset is not supported. (2) The endurance athlete or person about to do sustained heavy physical work in adverse conditions (marathon training block, expedition, military deployment, ski trip). The 50% incidence reduction in this subgroup is the strongest signal in the literature; 200 mg–1 g/day starting before the stress event is reasonable Hemilä & Chalker 2013. (3) Male history of kidney stones. The supplement should likely be avoided; food vitamin C is fine.

Alternatives

The honest comparator: zinc lozenges have similar-magnitude trial evidence (~1–2 day duration reduction at onset, formulation-dependent) and a different side-effect profile (taste, nausea, intranasal zinc can cause anosmia). Daily vitamin D in deficient adults has more consistent prevention signal in respiratory-tract-infection meta-analyses. Sleep duration (population-scale effect on infection susceptibility) and influenza/COVID vaccination dwarf any micronutrient effect. Vitamin C does not stand out as a uniquely effective cold intervention; it is one of several modest-effect tools.

Failure modes

The pattern that produces the "vitamin C doesn't work" reader experience: starting a single 1 g dose the day symptoms appear and concluding nothing happened. That is the exact protocol the trials show doesn't reliably work. The pattern that has trial support — start daily 500 mg–1 g months earlier and stay on it — is the one most lay users skip. Conversely, the chronic megadoser (3–5 g/day) who reports "I never get sick" is mostly anecdote (small effect, baseline rate, confounding); they may also be quietly setting themselves up for GI symptoms or, if male, kidney stones.

Practicalities

Cost is negligible: bulk ascorbic acid powder runs roughly $10–20/year at 500 mg–1 g/day. Standard 500 mg or 1 g tablets are sold everywhere. Effort is minimal — one tablet with food. Buffered or liposomal forms reduce stomach irritation at higher doses; they are not necessary at 500 mg–1 g. Effervescent "immune-boost" tablets dose vitamin C at the same range but carry added sugar / sodium and cost 10× more per gram. Food sources easily reach 100–200 mg/day: one bell pepper, one orange, a half-cup of broccoli — the RDA is genuinely reached by anyone eating fruit and vegetables.

History

The popular case for vitamin C against colds was made by Linus Pauling, twice Nobel laureate, in Pauling 1970 — he recommended 1–4 g/day and later escalated to far higher doses. The book sold millions, and "take vitamin C when you feel a cold coming on" entered the lay vocabulary on the strength of his authority. The medical establishment pushed back hard; Karlowski et al. 1975 (NIH/JAMA) was the influential negative trial that concluded vitamin C had no biological effect, though the blinding failure (placebo lactose distinguishable by taste) muddied that interpretation. The subsequent half-century of trials has landed between Pauling and his critics: there is a small biological effect, mostly on duration and severity with regular use; the magnitude is nowhere near Pauling's claims; and the at-onset protocol he popularised is the one the data least support.

Stakes / payoff

The honest projection at the population level is small. An adult who picks up routine 500 mg–1 g/day can expect, against their baseline ~2–3 colds per year, roughly 8% shorter episodes — about half a day off the typical 6–7 day cold. Severity-of-bad-days drops more sharply (around a quarter for the bedridden fraction). At the individual scale this is below detection in any given year; over a decade it is plausibly a week of fevered indoor days avoided. For the athlete or expedition member, the prevention effect is large enough to be felt — halving incidence on the trial scale — and the cost-benefit obviously favours taking it. For everyone else: a modest hedge, honestly framed.

Out of scope

This entry covers oral vitamin C for prevention and acute treatment of the common cold in the general adult population. It does not cover: IV vitamin C for sepsis or oncology indications (different mechanism, different evidence base); vitamin C for influenza-A, COVID-19, or pneumonia (separately reviewed in Hemilä 2017); scurvy treatment (clearly therapeutic at ≥30 mg/day); topical vitamin C for skin (separate substance, separate evidence). Dietary vitamin C status from food is mentioned only where supplementation is contrasted against it.

The credibility range

The optimist case. Vitamin C is biologically active in immune function (mechanism solid, plasma levels fall during infection); a large pooled trial base shows a real, statistically significant reduction in cold duration with regular use; severity reduction is larger than duration reduction and concentrated on the disabling end of the symptom spectrum (the 26% drop in severe-symptom days from Hemilä & Chalker 2023); a robust 50% incidence reduction exists in the physically-stressed subgroup; and the dose-response signal at 6–8 g/day in Anderson's trials hints that the optimal dose was never properly tested. The Karlowski blinding failure suggests the modal "vitamin C does nothing" narrative was over-stated; corrected analyses of the truly-blinded subgroup left a small genuine effect Karlowski et al. 1975. Side effects at ≤2 g/day are minimal in non-stone-formers, and it is one of the cheapest interventions in the catalogue. A reasonable optimist concludes: take 500 mg–1 g daily; the expected value is modestly positive and the floor is zero.

The skeptic case. The general-population incidence effect is null at the 11,306-participant level — the Cochrane RR upper bound is 1.00 Hemilä & Chalker 2013. The duration reduction is 8% in adults: clinically trivial — roughly half a day on a typical 6–7 day cold. Therapeutic dosing at symptom onset, the protocol most lay users actually follow, shows no consistent benefit. The severity-reduction signal is real but selectively re-cut from the same trial base, so the appearance of a "newer finding" is partly an artefact of analysis depth, not new data. Pauling's mega-dose claims have not held up; the at-onset Pauling protocol has the weakest support. High-dose supplements double kidney-stone risk in men Thomas et al. 2013, and the at-2 g/day UL exists because GI side effects are real. A reasonable skeptic concludes: dietary vitamin C is enough; supplements add nothing the population can measure, and the at-onset megadose ritual is folk medicine with documented downside.

The author's call. Land between the two, closer to the optimist on regular low-dose, closer to the skeptic on at-onset megadose. The evidence is high-quality (large Cochrane base, multiple meta-analyses, decades of trials — evidence 4) and the field is broadly aligned on the major findings, with continuing minor disagreement on dose-response and the at-onset signal (controversy 2). The intervention is honest-but-modest: routine 200 mg–1 g/day is a near-zero-cost hedge with a small expected benefit on cold duration and severity; the lay habit of grabbing a 1 g tablet at the first sniffle is the wrong protocol — benefit comes from prior regular use, not rescue dosing. The athlete / heavy-physical-stress case is the only one where the magnitude is clinically meaningful at the individual level. The action verb is do, cadence daily; the catch (kidney stones in male stone-formers) is real but narrow.

Stakeholder + incentive map

• Supplement industry. Vitamin C is one of the highest-volume supplements; effervescent "immune support" tablets (Emergen-C, Airborne) market the at-onset use case heavily despite the trial evidence being weakest there. Commercial incentive favours the rescue-dose narrative because it generates impulse purchases at the cold-season peak.
• Linus Pauling Institute (Oregon State). Institutional home of the pro-vitamin-C position; current institute guidance is more modest than Pauling's mega-dose claims but still recommends ~1 g/day with a 2 g/day ceiling and credits a real prevention-and-duration effect LPI 2018. Carries some legacy reputational interest in the substance.
• Mainstream nutrition / regulatory bodies. NIH ODS, USPSTF, Mayo Clinic frame routine vitamin C supplementation for cold prevention as not generally justified NIH ODS 2021. The institutional case is conservative: stick to the RDA via food, supplements unnecessary for most.
• Cochrane / Hemilä. Harri Hemilä (Helsinki) has produced most of the high-quality meta-analytic work; his framing is more sympathetic to vitamin C than the mainstream skeptical reading — he tends to emphasise the severity effect, the dose-response, and the Karlowski blinding flaw. Not commercial; a real intellectual investment in the substance.
• Lay culture. Orange juice, citrus folk wisdom, and the orthomolecular movement all reinforce the at-onset megadose. Strong cultural priors; weak evidentiary fit.

Population variability

• Baseline vitamin C status. Adults already at biochemical sufficiency (most people eating any fruit/vegetables) have less marginal benefit; the trial population is mostly sufficient. Subclinically deficient subgroups (heavy smokers, very-low-vegetable diets) plausibly respond more, though this is not the modal trial population.
• Children vs adults. Children show a larger duration reduction (14% vs 8%) and a larger dose-response (1–2 g/day shortens by 18%) Hemilä & Chalker 2013. Trial doses in children are typically capped at 1–2 g/day and the GI ceiling is lower.
• Physical-stress subgroup. The 50% incidence reduction in marathon runners, skiers, and soldiers on subarctic exercises is the largest signal in the field. Mechanism is plausible (acute stress depletes vitamin C; the stressed cohort starts at the deficit edge). Generalisation: planning a heavy physical block in cold conditions is the one case where the prevention effect is large enough to matter clinically.
• Smokers. Higher RDA (additional 35 mg/day) on biochemical grounds NIH ODS 2021; not a separate cold-prevention case.
• Male stone-formers. Doubled kidney-stone risk on supplemental high-dose ascorbic acid Thomas et al. 2013. The cohort was male; female evidence is thinner.
• Hereditary hemochromatosis. Vitamin C enhances non-heme iron absorption; high-dose supplementation worsens iron overload. Dietary vitamin C is not implicated at the magnitude that matters.

Knowledge gaps

• Optimal at-onset dose. Anderson's trials (6 and 8 g/day on day 1) showed continued dose-response at the highest dose tested. No trial has formally tested 10–16 g/day at onset; the curve plateau is unknown Hemilä 2017. Doing so safely is hard — GI side effects at those doses are nearly universal.
• Why does the physical-stress subgroup respond so much more? The 50% effect is consistent across five disparate trials but the mechanism — whether it is stress-induced vitamin C depletion, increased oxidative load, immune modulation by acute exercise, or selection of a cohort with marginal baseline status — is not nailed down.
• Severity vs duration. The 2023 severity meta-analysis is a re-cut of existing trials, not a new prospective design. A trial purpose-built to discriminate "fewer bedridden days" from "fewer total days" would settle whether the severity effect is the real signal and the duration effect a confounded version of it.
• Real-world adherence vs trial adherence. Trial-level regular use is daily, compliance-monitored. Lay regular use is patchy. The "real-world equivalent" of the trial-arm benefit is plausibly smaller than the trial effect.
• Female kidney-stone risk. The Thomas cohort was male; women's stone-formation response to high-dose supplemental vitamin C is less well characterised Thomas et al. 2013.

Scope alignment with brief. The brief named four consequences: cold incidence, symptom duration, symptom severity, the prophylactic-vs-acute contrast, and the tolerable upper intake. All five are covered end-to-end. The prophylactic-vs-acute contrast is the spine of the article rather than a side note — it is the single biggest source of reader misuse, and the trial evidence on the two patterns diverges sharply enough that subordinating it to a sub-section would have buried the lede.

Score calls. A few were close:

• health_short_term at 2 rather than 3: the duration reduction is real but small (~half a day per cold) and the severity effect, while larger in proportion, only matters on the bedridden fraction. A 3 ("clear functional improvement") would overstate what the typical adult notices.
• longevity at 0: no credible mortality or major-disease-prevention signal for vitamin C supplementation at these doses in the general population. Cold-prevention does not aggregate to longevity at any meaningful effect size.
• energy at 0: the "less time sick" path is real but indirect and below the threshold for a dimension score; scoring it would let zinc, vitamin D, sleep, and several other interventions also accumulate phantom energy points by the same logic.
• evidence at 4 rather than 5: Cochrane base is at the 5 ceiling for breadth, but the field has genuine remaining disagreement on therapeutic dose-response and on the magnitude of the severity-vs-duration split. Holding back from 5 preserves the rule that 5 means "no meaningful dispute."
• controversy at 2: the field has settled on the major findings; what remains (at-onset signal, Karlowski reinterpretation, severity meta-analysis cuts) is minor pushback at the margins, not foundational disagreement.
• applicability at 5: per the meta spec's avoidance / decision-audience rule, the addressable audience is everyone weighing the most-popularised cold supplement, not just current users.
• pull at 2: borderline 2/1. The daily-tablet ritual feels like nothing; the lay at-onset megadose ritual feels active and protective but is exactly the pattern the article tells the reader to drop. Net pull is neutral.

Dream narrative call. Overall score works out below 40, so the narrative was optional. Wrote one anyway on the relief / debunking lever — the reader's payoff is not being misled by the popular protocol — because the entry's hook is genuinely a re-framing of a widely-held wrong belief, and the dek and tagline lean on that hinge.

Therapeutic-dose / Anderson trials. Cited inside the dossier but left out of the article body proper. The two-trial dose-response signal at 6–8 g/day is interesting and partially justifies further research, but reader-facing inclusion would muddy the central misconception the article is trying to correct ("at-onset megadose doesn't work, so don't do it"). The honest reader-facing position is: pooled data shows no consistent at-onset benefit; the dose-response hint is a research footnote, not an action.

Retracted Ran 2018 meta-analysis. Mentioned in the dossier with the retraction noted; deliberately not cited in the article body. The corrected effect estimates lost significance, so there is nothing the reader needs to act on, and citing a retracted paper just to dismiss it adds noise.

Future links. Three sibling entries this should cross-link to once they exist:

• Zinc for the common cold — the natural at-onset comparator with the inverted evidence pattern (real acute signal, no prevention signal).
• Vitamin D for respiratory infections — the modest-effect prevention neighbour with a better deficiency-corrected signal.
• Sleep and infection susceptibility — the dominant behavioural lever the alternatives section names; vitamin C should defer to it.

Out-of-scope and separate-entry candidates. IV vitamin C for sepsis and oncology adjunct work is large enough and clinically distinct enough to warrant its own entry; flagged in the body but does not belong here. The same applies to vitamin C in COVID-19, which has its own evidence base. Both are separate substances in clinical effect, even though the molecule is the same.

What stayed out. Liposomal / buffered / "natural" vitamin C marketing distinctions are mentioned briefly in practicalities but not given their own section — the answer is "no meaningful clinical edge for cold outcomes" and that resolves cleanly in one paragraph. Histamine / antihistamine effects of vitamin C are real at very high doses but irrelevant to cold outcomes and would dilute the article.