Every generation of clinicians inherits a remedy that seems to promise more than mere biology should allow. For our forebears at sea it was citrus, the difference between life and death on a latish eighteenth‑century deck. For late‑twentieth‑century dreamers it was Linus Pauling’s crystalline faith in ascorbate. And for many of us in critical care over the last decade, vitamin C became that rare thing—an old molecule with a new mystique. This is the love story of an idea, told with trials, tempering, and the tug of evidence.
Act I — Sailors, Scurvy, and the First “Randomised” Trial
James Lind’s famous 1747 comparison of citrus versus other cures for scurvy is often cast as the first randomised clinical trial, though the assignment was not truly randomised in a modern sense. Still, Lind’s method—comparing like with like, varying one thing—was revolutionary. His sailors receiving oranges and lemons recovered rapidly; those given cider, vinegar, seawater, or elixirs did not. The experiment seeded a clinical tradition: disciplined comparison over anecdote, physiology tempered by proof.1
Act II — Pauling’s Promise
Two centuries later, double Nobel laureate Linus Pauling argued that vitamin C could prevent colds and improve well‑being, catalysing a popular fascination that has barely dimmed. Pauling’s claims outran the data, but he was right about one thing: ascorbate biology is astonishingly rich, extending from co‑factor roles in catecholamine synthesis to redox buffering under stress.2 Interest spilled into oncology, where the National Cancer Institute still maintains a sober review of intravenous vitamin C as adjunct therapy—promising hypotheses, mixed human data, unresolved questions of dose and delivery.3
Critical care would soon claim its own chapter. Ascorbate participates in dopamine β‑hydroxylase–mediated conversion of dopamine to noradrenaline, supports endothelial function, and scavenges reactive species—physiological hooks tailor‑made for the catecholamine‑hungry, oxidant‑rich milieu of septic shock.4
Act III — The Spark: an ICU Before–After and the CCR Stage
In early 2017, Marik and colleagues published a single‑centre before–after study in Chest (47 patients per epoch) suggesting that hydrocortisone, vitamin C, and thiamine dramatically reduced hospital mortality in severe sepsis/septic shock (8.5% vs 40.4%; adjusted odds of death 0.13, 95% CI 0.04–0.48).5 The story then stepped onto the Critical Care Reviews stage in Belfast, where Marik’s 2017 CCR talk captured global imagination and scepticism in equal measure.6
Within two years, Fowler et al. reported CITRIS‑ALI, a multicentre RCT of septic patients with ARDS randomised to high‑dose vitamin C (50 mg/kg every 6 h for 96 h) or placebo. The three primary endpoints—change in SOFA, CRP, and thrombomodulin at 96 h—were negative. A nominally lower 28‑day mortality emerged among those receiving vitamin C (secondary outcome), but the trial was not adjusted for multiplicity; the authors correctly counselled caution.7
Then came randomisation of the Marik “cocktail” itself. The VITAMINS trial, led by Fujii and colleagues and presented live57 at CCR20, compared hydrocortisone+vitamin C+thiamine to hydrocortisone alone in septic shock. The primary outcome—hours alive and vasopressor‑free at 7 days—did not differ (median 122.1 vs 124.6 h; difference −2.5 h, 95% CI −10.0 to 5.1).8 In the same era, ACTS (n=200) found no reduction in change in SOFA at 72 h (between‑group difference −0.8, 95% CI −1.7 to 0.2) and a neutral signal for mortality (HR 1.3, 95% CI 0.8–2.2).9 VICTAS (n=501) was halted for funding/time, with no difference in ventilator‑/vasopressor‑free days at 30 days (median 25 vs 26; difference −0.2 days, 95% CI −3.5 to 3.8).10 ATESS, a meticulous Korean RCT, similarly returned neutral findings.11
Finally, the LOVIT trial, reported in NEJM, asked a sobering question: might vitamin C be harmful in the sickest? In 872 adults with sepsis on vasopressors, high‑dose vitamin C (50 mg/kg q6h for 96 h) increased the composite of death or persistent organ dysfunction at day 28 compared with placebo (44.5% vs 38.5%; absolute difference 6.0 percentage points, 95% CI 0.8–11.2; RR 1.21, 95% CI 1.04–1.40).12
By 2023, a careful meta‑analysis of intravenous vitamin C monotherapy in critically ill adults concluded that pooled effects on mortality were null, with heterogeneity but no convincing signal to treat outside trials.13
An important aside on controversy
In March 2022, media coverage reported statistical allegations that the original 2017 before–after study might rely on fraudulent data; these were not peer‑reviewed, were contested by the authors, and—critically—the Chest paper remains published without retraction or an expression of concern.14,5 Whatever one’s view, the allegation episode further emphasised the need to privilege neutral, well‑conducted RCTs over dramatic observational findings.
Interlude — The Antioxidant Paradox: When Vitamin C Turns Pro‑oxidant
It is tempting to focus only on potential benefit when an intervention is physiologically sensible. But redox biology is not a one‑way street, and so too for ascorbate. At pharmacological plasma concentrations achievable only by intravenous infusion, ascorbate can reduce transition metals (e.g., Fe3+→Fe2+), catalysing Fenton chemistry and generating hydrogen peroxide (H2O2) in extracellular fluid; elegant in vivo work in animals and humans has demonstrated measurable ascorbate radicals and H2O2 generation under these conditions.15 Contemporary reviews likewise frame ascorbate as a compound with “two faces”—antioxidant at physiological concentrations, potentially pro‑oxidant at millimolar levels depending on catalytic metal availability and compartmentalisation.16
Why does this matter in the ICU? Because sepsis is an iron‑rich, inflamed, dysregulated state. A therapy capable of donating electrons and spawning H2O2 might—counter‑intuitively—exacerbate oxidative injury or alter microvascular signalling in susceptible tissues. Add practical risks such as oxalate nephropathy (vitamin C is metabolised to oxalate) and analytical artefact (interference with some point‑of‑care glucose meters), and the case for balance is compelling. Biopsy‑proven oxalate nephropathy has been described after high‑dose intravenous vitamin C in severely ill patients, including two COVID‑19 cases (AKI with calcium oxalate deposition) and additional series during the pandemic period.17,18 High ascorbate levels can also produce spurious hyperglycaemia on certain glucose dehydrogenase–based point‑of‑care meters, particularly in renal impairment—risking inappropriate insulin and hypoglycaemia if laboratories are not used for confirmation.19,20
Therapeutic deliberation demands symmetry: opportunity is necessary but not sufficient; one must weigh plausible harm with equal seriousness before embracing a therapy at scale. Vitamin C embodies that duty of balance.
Mechanisms of Action — How Might Vitamin C Work?
If our romance with vitamin C began with sailors and was rekindled by sepsis, the molecular script is no less dramatic. In endothelium—the organ that lines the vascular tree—ascorbate concentrates via SVCT transporters, augments nitric oxide signalling, and stabilises the permeability barrier under inflammatory stress.21 One well‑defined route is through tetrahydrobiopterin (BH4) biology: ascorbate increases intracellular BH4, recoupling endothelial nitric oxide synthase (eNOS) and restoring physiologic NO generation when oxidative stress would otherwise tilt the enzyme towards superoxide.22,23,24
Vitamin C is also a classical co‑factor for several Fe2+/2‑oxoglutarate–dependent hydroxylases. That includes prolyl and lysyl hydroxylases in collagen biosynthesis—basic to vascular integrity and wound strength—and the two carnitine biosynthetic hydroxylases that help sustain mitochondrial fatty‑acid entry during stress.25,26 At the neurohumoral interface, ascorbate enables peptidylglycine α‑amidating monooxygenase (PAM) to convert glycine‑extended precursors into amidated hormones (e.g., vasopressin‑family peptides), and supports catecholamine synthesis via dopamine β‑hydroxylase—mechanisms that sit naturally with shock physiology.27,4
Beyond classical enzymology, ascorbate modulates hypoxia and gene regulation. As a reducing co‑factor for prolyl‑hydroxylases that target HIF‑α, ascorbate promotes HIF degradation and may temper hypoxia‑inflammation crosstalk; it also enhances TET/JmjC dioxygenases that drive active DNA and histone demethylation—pathways reported to be ascorbate‑sensitive in immune and epithelial cells.28,29
Immunologically, vitamin C accumulates in neutrophils, supports chemotaxis and phagocytosis, and may modulate NET formation; deficiency impairs several of these functions in vivo.30 At the microvascular level, animal and cellular models show improved capillary perfusion, reduced leak, and preservation of junctional proteins during inflammatory stress—effects that map to inhibition of NADPH oxidase signalling and protection of tight‑junction phosphorylation states.31,32,33,34