Vitamin D and viral infection: what 164 studies do and do not show - review

The Modulatory Role of Vitamin D in the Molecular Mechanisms of Sepsis and Infections with Focus on Viral Pathogenesis: A Narrative Review

Pathogens (MDPI), 25 August 2026, https://doi.org/10.3390/pathogens15090894

Federica Vincenzi, Maria Grazia Crobu, Stelvio Tonello, Nicole Vercellino, Elena Grossini, Paolo Ravanini, Rosalba Minisini, Lucio Boglione, Mario Pirisi, Pier Paolo Sainaghi, Carlo Smirne

Summary by Claude - August, 2026

An Italian group at Università del Piemonte Orientale pulls together the mechanistic and clinical literature on vitamin D in viral infection, and then does something unusual — it states plainly that the link to viral sepsis is largely assumed rather than demonstrated: "direct evidence linking vitamin D to viral sepsis specifically remains particularly limited, with most mechanistic inference extrapolated from non-septic viral infection and bacterial sepsis literature."

This is a narrative review (not systematic, no pooled data, ~164 references). It uses 30–150 ng/mL as sufficiency, 21–29 insufficiency, under 20 deficient.

The mechanism sections are the strongest part: calcitriol directly induces cathelicidin (LL37) through a VDRE in the gene promoter; defensin-β2 needs both VDR and NF-κB/IL-1β signaling; calcitriol suppresses hepcidin, freeing ferroportin to export iron and starving pathogens of it; it drives autophagy (shown in HIV-infected macrophages and influenza-infected A549 cells); it builds tight junctions (occludin, claudins); it raises Nrf2 and calms oxidative stress; and single-cell RNA-seq shows it boosting RIG-I/MDA5 and JAK/STAT while shifting Th1 cells toward regulatory phenotypes.

Two summary tables collect the clinical trials with doses. Highlights: Martineau's 25-RCT, 10,933-participant meta-analysis found daily or weekly dosing protective against acute respiratory infection while bolus dosing was not, with the greatest benefit in deficient people; 1,200 IU/day for 90 days cut influenza A in schoolchildren; 40,000–50,000 IU within 72 hours of COVID diagnosis cut 14-day mortality; a single 200,000 IU dose in hospitalized COVID patients did nothing. In HCV, 2,000 IU/day improved sustained virologic response — but the authors note this is now historical, since direct-acting antivirals already cure ≥96%.

Summary by type of virus

Category Refs What the evidence actually is
SARS-CoV-2 / COVID-19 ~25 The largest block by far — 4 RCTs, several observational cohorts, 1 meta-analysis, rest mechanistic/narrative
HIV 12 2 RCTs, 1 meta-analysis, cohort + in-vitro autophagy work
RSV 9 2 birth cohorts, 1 severity cohort, VDR-SNP meta-analysis, cell-barrier mechanism
Influenza 6 2 RCTs (both Urashima), 1 meta-analysis of 10 RCTs, autophagy + peptide mechanism
HBV 4 All observational or in-vitro — no RCT
HCV 4 2 interventional (both PEG-IFN era), 1 cohort, 1 oxidative-stress study
EBV 3 1 supplementation trial in MS patients, 2 hypothesis/review
HSV 1 Cathelicidin-derived peptide, in vitro only
Dengue shared 1 Mentioned only inside a broader review title
Mixed acute respiratory infection / pneumonia (pathogen not separated) ~15 Includes the two big Martineau/Jolliffe meta-analyses and 4 pediatric pneumonia RCTs
Sepsis — bacterial or unspecified + critical care ~26 Every sepsis intervention trial cited (VITdAL-ICU, VIOLET, Leaf, Quraishi)
Viral sepsis specifically 3 Two definition/epidemiology papers and one CAP cohort — no trial
Vitamin D biology, immunology, transcriptomics, non-infectious ~56 Background and mechanism

That last grouping is the paper's own headline finding: 26 sepsis studies, none of them viral, against 3 viral-sepsis references that contain no intervention.

Benefits found by Virus

Acute respiratory infection (mixed viruses) is the strongest signal. Martineau's individual-participant meta-analysis of 25 RCTs and 10,933 people found daily or weekly dosing protective, bolus dosing not protective, and the effect concentrated in people who were deficient to start with. This remains the single most defensible claim in the paper.

Influenza has two positive pediatric RCTs. 1,200 IU/day for 90 days reduced influenza A in schoolchildren (no effect on influenza B). 2,000 IU/day reduced infection in the first month only, with no overall effect at two months — which reads more like a seasonal-timing artifact than a durable benefit. A 10-RCT meta-analysis of 8,029 patients found reduced influenza risk overall.

COVID-19 is where the trials disagree most sharply, and the pattern is about timing and form. Calcifediol (0.532 mg then 0.266 mg) alongside standard care cut ICU admission. 40,000–50,000 IU given within 72 hours of infection cut 14-day mortality regardless of baseline level. 60,000 IU sped viral clearance in deficient patients. But a single 200,000 IU dose in already-hospitalized patients did nothing to length of stay, ICU admission or mortality. Early and repeated beat late and huge.

RSV benefit is entirely prenatal and observational — low cord-blood levels predict bronchiolitis in the first year, and low admission levels track with severity. There is no RSV supplementation RCT here. The FokI (rs2228570) VDR polymorphism is associated with the most severe cases, which is a genuinely useful lead.

HIV shows the clearest immunologic benefit: 4,000–7,000 IU/day for 12 weeks raised CD4+ counts and slightly lowered viral load; 120,000 IU/month for 12 months reduced activated CD4+, CD8+ and monocyte fractions. A meta-analysis of 10 RCTs found no CD4+ effect, so this is unsettled.

HBV has mechanism but no trial — calcitriol suppresses the HBV core promoter in hepatocytes, and low levels track with high viral replication, but causal direction is unresolved.

HCV benefit (2,000 IU/day improving sustained virologic response) is real but the authors correctly retire it: direct-acting antivirals already cure ≥96%, so there's nothing left to add.

EBV supplementation lowered anti-EBNA-1 IgG but did not improve cytotoxic CD8+ response — a biomarker moved, an outcome didn't.

PDF

Related in VitaminDWiki

COVID

HIV

Herpes (HSV)

RSV

Epstein-Barr Virus