Vitamin C and BV: treatment and recurrence are different questions.

First, the clinical context.

Bacterial vaginosis involves a change in the vaginal bacterial community. The CDC recommends treatment for symptomatic BV and lists antibiotic regimens; vaginal vitamin C is not listed among its recommended regimens. CDC treatment guidance: Centers for Disease Control and Prevention ↗

This guide explains research. It does not recommend substituting vitamin C for prescribed treatment. Use new or recurring symptoms as a reason to seek clinical evaluation.

The 2011 treatment trial.

Among 277 randomized women with BV, the reported clinical cure proportions were 55.3% with vitamin C and 25.7% with placebo. These are the intention-to-treat results, a useful starting point because they retain the randomized comparison. Original 2011 paper: Petersen and colleagues ↗

The widely repeated, higher subgroup figure answers a narrower question. Selecting that figure without explaining who was included changes the impression of the trial. Short follow-up and manufacturer-affiliated authors are also relevant context.

The 2013 recurrence trial.

This trial enrolled 142 women after successful antibiotic treatment. At six months, recurrence was 16.2% with vitamin C versus 32.4% with placebo. The earlier, three-month comparison was not statistically significant. Original 2013 paper: Krasnopolsky and colleagues ↗

The difference at six months is 16.2 percentage points; the relative reduction is 50%. Both describe the same comparison. It was a manufacturer-funded trial of repeated monthly courses, not a test of avoiding antibiotics during an active episode.

An encouraging result still needs a confidence check.

Earlier work included a 100-participant trial using an older “non-specific vaginitis” definition. Diagnostic criteria, follow-up and participant selection vary across this literature. 2004 trial: Petersen & Magnani ↗

The Khaikin review rated key comparisons as low or very low certainty. Pooling studies can clarify a signal, but does not erase problems in the underlying evidence. 2026 synthesis: Khaikin and colleagues ↗

Look for replication, longer follow-up and patient-important outcomes when reading future work. Our library separates trials from reviews so the same participants are not accidentally counted twice.