Evidence Trace, complete trial record

In plain terms

In plain terms: Vitamin C (ascorbic acid) has some human trial evidence supporting it, but the evidence is not yet strong enough for EBL to call it established. It is worth knowing about and worth discussing with your clinician.

Vitamin C (ascorbic acid)

B Supported: a good trial, backed by more

Category Supplements

Evidence Pooled reviews of randomized trials and large randomized trials

Reviewed October 2026

PMIDs 23440782 · 38082300 · 18997197 · 17698683 · 19066368 · 19116389 · 35704292 · 42267875

The endpoint this grade is based on: self_report, Slightly shorter and milder colds when vitamin C is taken every day before a cold starts, at 200 mg a day or more in the trials; about 8 percent shorter in adults and 14 percent in children. It did not prevent colds in the general population, and starting it once a cold began had no consistent effect.

The evidence as the record states it

Evidence record

PMID 23440782 PMID 38082300 PMID 18997197 PMID 17698683 PMID 19066368 PMID 19116389 PMID 35704292 PMID 42267875

Hemilä and Chalker (Cochrane Database Syst Rev, 2013; PMID 23440782): placebo-controlled trials of at least 200 mg a day; daily use shortened colds by 8 percent in adults and 14 percent in children, did not prevent colds in the general community (risk ratio 0.97), halved colds in 598 people under heavy physical stress, and starting vitamin C after symptoms began had no consistent effect. Hemilä and Chalker (BMC Public Health, 2023; PMID 38082300): colds 15 percent less severe at 1 g a day or more. Physicians' Health Study II, Sesso et al. (JAMA, 2008; PMID 18997197) and Gaziano et al. (JAMA, 2009; PMID 19066368): 14,641 men, 500 mg a day for a mean 8 years; no effect on major cardiovascular events, cancer or total mortality. Women's Antioxidant Cardiovascular Study, Cook et al. (Arch Intern Med, 2007; PMID 17698683) and Lin et al. (J Natl Cancer Inst, 2009; PMID 19116389): women at high risk, 500 mg a day for a mean 9.4 years; no effect on cardiovascular events or cancer. LOVIT, Lamontagne et al. (N Engl J Med, 2022; PMID 35704292): intravenous vitamin C in sepsis raised death or persistent organ dysfunction (risk ratio 1.21). VICTORY, Stoppe et al. (JAMA, 2026; PMID 42267875): intravenous vitamin C in severe burns, stopped early, 28-day death 15.0 against 7.6 percent. Grade B for cold duration with daily use rests on many independent trials of a self-reported outcome; the other uses carry their own letters.

PMID 23440782 PMID 38082300 PMID 18997197 PMID 17698683 PMID 19066368 PMID 19116389 PMID 35704292 PMID 42267875

Evidence by specific use

Common-cold duration with daily supplementation

Grade B

Hemilä and Chalker (Cochrane Database Syst Rev, 2013; PMID 23440782) pooled 31 placebo-controlled comparisons with 9,745 cold episodes: taking vitamin C every day shortened colds by 8 percent in adults (95% CI 3 to 12) and 14 percent in children (7 to 21), roughly half a day on a week-long cold in an adult. Hemilä and Chalker (BMC Public Health, 2023; PMID 38082300) pooled 15 comparisons from 10 randomized, double-blind trials at 1 g a day or more and found colds 15 percent less severe (9 to 21). Many independent trials, consistent in direction, on a self-reported endpoint, which caps the letter at B. The effect is small, and it applies only to vitamin C taken every day before the cold starts.

PMID 23440782 PMID 38082300

Shortening a cold by starting vitamin C after symptoms begin

Grade D

Hemilä and Chalker (Cochrane Database Syst Rev, 2013; PMID 23440782): in the 7 comparisons where vitamin C was started after the cold began, there was no consistent effect on how long the cold lasted or how bad it was. Tested, no benefit.

PMID 23440782

Preventing colds in the general population

Grade D

Hemilä and Chalker (Cochrane Database Syst Rev, 2013; PMID 23440782): in general community trials with 10,708 participants, the chance of catching a cold was the same on vitamin C and placebo (risk ratio 0.97, 95% CI 0.94 to 1.00). Tested, no benefit.

PMID 23440782

Preventing colds during heavy physical stress

Grade C

Hemilä and Chalker (Cochrane Database Syst Rev, 2013; PMID 23440782): in 5 trials with 598 marathon runners, skiers and soldiers on subarctic exercises, colds were about half as common on vitamin C (risk ratio 0.48, 95% CI 0.35 to 0.64). The trials are small and the outcome is self-reported, so the letter stays at C.

PMID 23440782

Major cardiovascular events

Grade D

Physicians' Health Study II (Sesso et al., JAMA, 2008; PMID 18997197): 14,641 men, 500 mg a day for a mean 8 years; major cardiovascular events unchanged (hazard ratio 0.99, 0.89 to 1.11), and total mortality unchanged (1.07, 0.97 to 1.18). Women's Antioxidant Cardiovascular Study (Cook et al., Arch Intern Med, 2007; PMID 17698683): 8,171 women at high risk, 500 mg a day for a mean 9.4 years; no effect on the combined outcome (relative risk 1.02, 0.92 to 1.13). Two large trials in men and women, both null: tested, no benefit.

PMID 18997197 PMID 17698683

Cancer prevention

Grade D

Physicians' Health Study II (Gaziano et al., JAMA, 2009; PMID 19066368): 14,641 men, 500 mg a day for a mean 8 years; total cancer (hazard ratio 1.01, 0.92 to 1.10) and prostate cancer (1.02, 0.90 to 1.15) unchanged. Women's Antioxidant Cardiovascular Study (Lin et al., J Natl Cancer Inst, 2009; PMID 19116389): 7,627 women, 500 mg a day for an average 9.4 years; total cancer relative risk 1.11 (0.95 to 1.30) and cancer death 1.28 (0.95 to 1.73), neither significant. Two large trials, both null: tested, no benefit.

PMID 19066368 PMID 19116389

Blood pressure

Grade C

Juraschek et al. (Am J Clin Nutr, 2012; PMID 22492364) pooled 29 randomized trials, median 500 mg a day for a median 8 weeks, each with 10 to 120 participants: systolic pressure fell 3.84 mmHg (95% CI 2.38 to 5.29) and diastolic 1.48 mmHg. Kosari et al. (Food Funct, 2026; PMID 41784995) re-pooled 17 trials and found about 3.7 mmHg lower systolic pressure, rating the certainty low. Blood pressure is an established stand-in for heart attack and stroke, but every trial is small and short, which holds the letter at C.

PMID 22492364 PMID 41784995

New gout diagnoses

Grade C

Juraschek et al. (Am J Clin Nutr, 2022; PMID 35575611) went back to the Physicians' Health Study II after it ended: men assigned 500 mg a day reported 12 percent fewer new gout diagnoses (hazard ratio 0.88, 0.77 to 0.99). It was an analysis added after the trial, on a self-reported diagnosis, in one trial. In people who already had gout, Stamp et al. (Arthritis Rheum, 2013; PMID 23681955) found 500 mg a day for 8 weeks lowered uric acid by only 0.23 mg/dL, against 1.9 mg/dL for allopurinol. One positive analysis nobody has repeated: C.

PMID 35575611 PMID 23681955

Preventing complex regional pain syndrome after a wrist fracture

Grade C

Zollinger et al. (Lancet, 1999; PMID 10636366): 123 adults with wrist fractures, 500 mg a day for 50 days; the pain syndrome then called reflex sympathetic dystrophy occurred in 7 percent of wrists on vitamin C against 22 percent on placebo. Zollinger et al. (J Bone Joint Surg Am, 2007; PMID 17606778), the same group: 416 patients, 2.4 percent against 10.1 percent, with 500 and 1,500 mg a day both effective. Ekrol et al. (J Bone Joint Surg Am, 2014; PMID 25187584), an independent group: 336 patients, 500 mg a day for 50 days, no difference in the syndrome, function or healing at one year. Evaniew et al. (J Orthop Trauma, 2015; PMID 26197022) pooled the three trials, 890 patients: risk ratio 0.45 (0.18 to 1.13), not significant, with low-quality evidence. Positive in one group's trials, null when an independent group repeated it: C.

PMID 10636366 PMID 17606778 PMID 25187584 PMID 26197022

Intravenous high-dose vitamin C in sepsis

Grade D

LOVIT (Lamontagne et al., N Engl J Med, 2022; PMID 35704292): 872 adults with sepsis on vasopressors; death or persistent organ dysfunction at 28 days in 44.5 percent on vitamin C against 38.5 percent on placebo (risk ratio 1.21, 1.04 to 1.40). CITRIS-ALI (Fowler et al., JAMA, 2019; PMID 31573637): 167 adults with sepsis and lung failure; no difference in the organ failure score. C-EASIE (Vandervelden et al., Crit Care, 2025; PMID 40269974): 300 emergency department patients, 1.5 g every 6 hours for 4 days; organ failure 8.7 percent lower, not significant. The largest trial found harm, the others found no benefit: tested, harm found. This is a hospital treatment and says nothing about vitamin C tablets.

PMID 35704292 PMID 31573637 PMID 40269974

Intravenous high-dose vitamin C in severe burns

Grade D

VICTORY (Stoppe et al., JAMA, 2026; PMID 42267875): 238 adults with severe burns. Death or persistent organ dysfunction at 28 days occurred in 40.8 percent on vitamin C against 29.7 percent on placebo (risk ratio 1.28, 0.99 to 1.65, P = .06), which crossed the trial's stopping threshold. Death within 28 days was higher on vitamin C, 15.0 against 7.6 percent (risk ratio 1.96, 1.32 to 2.90), and hospital death 23.3 against 16.1 percent. A hospital treatment, tested and possibly harmful; it says nothing about vitamin C tablets.

PMID 42267875

Why Grade B and not Grade A

Grade B reflects moderate human evidence -- typically a single strong trial, multiple smaller trials, or surrogate endpoint data without hard clinical outcome replication. To reach Grade A, Vitamin C (ascorbic acid) would require at least one additional independent randomized trial demonstrating a hard clinical endpoint or objectively measured functional outcome in a comparable population.

What this grade covers and what it does not

Grade B on a self-reported endpoint, for common-cold duration with daily supplementation. Hemilä and Chalker (Cochrane Database Syst Rev, 2013; PMID 23440782) pooled 31 placebo-controlled comparisons with 9,745 cold episodes: colds were 8 percent shorter in adults (95% CI 3 to 12) and 14 percent shorter in children (7 to 21) when vitamin C was taken every day before the cold began. A second pooling of 10 randomized, double-blind trials at 1 g a day or more found colds 15 percent less severe (Hemilä and Chalker, BMC Public Health, 2023; PMID 38082300). The trials come from many independent groups and agree in direction; cold duration is reported by the participants, which caps the letter at B. The effect is small. The same review found no prevention of colds in the general population (risk ratio 0.97) and no consistent effect when vitamin C was started after symptoms began; both are graded D below. WHAT THIS GRADE DOES NOT COVER: heart disease and cancer, tested without benefit in two large trials; blood pressure, gout and wrist-fracture pain syndrome, each graded C; intravenous vitamin C in hospital, which is a different treatment and was harmful in the largest sepsis trial and possibly in severe burns; and correction of scurvy or documented deficiency.

What would move this grade

An independent randomized trial demonstrating a hard clinical endpoint -- mortality, major adverse event, or objectively confirmed functional improvement -- in an adequately powered sample, with results independently replicated, would support Grade A.

How to track whether this is working No laboratory marker tracks the effect graded here; the trials counted cold days and symptoms in participant diaries.