How Many mg of Cranberry PACs Do You Actually Need?

Almost every article about cranberry repeats the same number: 36 mg of proanthocyanidins — a figure that does not appear in the NIH NCCIH consumer summary on cranberry at all. Very few explain where it came from, what happens on either side of it, or why the figure is impossible to check on most products sold today. This page is only about the amount. If you want the separate question of whether cranberry does anything at all, we covered the trial evidence in our review of cranberry and Lactobacillus for women's urinary health.

The 36 mg threshold, and the study that drew the line

The threshold is not a marketing round number. It comes out of a 2024 systematic review and meta-analysis in Frontiers in Nutrition by Xiong and colleagues, which pooled 10 randomised controlled trials covering 2,438 participants — 1,217 taking a cranberry product and 1,221 controls — and then split them by how much PAC the product actually delivered per day.

The split produced a clean break. Trials delivering at least 36 mg of PAC per day showed a statistically significant 18% reduction in risk. Trials delivering less than that showed nothing that reached significance. Same botanical, same outcome measure, different dose, opposite conclusion — which is a large part of why cranberry has looked so inconsistent in the literature for thirty years.

What the 2024 meta-analysis found, by subgroup

Subgroup	Risk ratio	95% CI	p	Significant?
PAC intake ≥36 mg/day	0.82	0.69 – 0.98	0.03	Yes — 18% lower risk
PAC intake <36 mg/day	—	—	0.39	No
Female-only cohorts	0.84	0.71 – 0.98	0.02	Yes
Mixed-gender cohorts	—	—	—	No
Use for 12 – 24 weeks	0.75	0.61 – 0.91	0.004	Yes
Use for under 12 weeks	—	—	—	No
Use for over 24 weeks	—	—	—	No

Source: Xiong Z, Gao Y, Yuan C, Jian Z, Wei X. Frontiers in Nutrition 2024;11:1422121. PMID 39668896.

Two things in that table deserve more attention than they usually get. The mixed-gender row is a genuine null result and it is rarely quoted. And the duration rows show that dose alone is not the whole story — the pooled benefit only reached significance in the trials that ran for 12 to 24 weeks. A correctly dosed product taken for three weeks is not a tested scenario.

The dose-response data: 18 mg vs 36 mg vs 72 mg

The meta-analysis tells you where the line is. It does not tell you what happens across a day. For that you need the dose-ranging study everyone cites and almost nobody reproduces: Howell and colleagues, published in BMC Infectious Diseases in 2010.

The design matters, so here it is in full. Thirty-two female volunteers over the age of 18, sexually active, with normal renal function, were recruited through four urology departments — eight each in Japan, Hungary, Spain and France — for a randomised, double-blind, placebo-controlled cross-over study. The volunteers in Japan and Hungary received 0, 36 or 72 mg of PAC equivalents per day; those in Spain and France received 0, 18 or 36 mg. Everyone took two capsules at 8:00 AM. Urine was pooled from 9:00 AM to 2:00 PM (the “1–6 h” sample) and again the following morning (the “24 h” sample). Anti-adhesion activity was measured with a mannose-resistant haemagglutination assay against uropathogenic E. coli.

Anti-adhesion activity in urine, by PAC dose

PAC dose (single morning dose)	Anti-adhesion activity at 1–6 h (median)	at 24 h (median)	Study	n receiving this arm
Placebo (0 mg)	0%	0%	Howell 2010	32
18 mg PAC	50%	0%	Howell 2010	16 (Spain, France)
36 mg PAC	90.6%	12.5%	Howell 2010	32 (all four sites)
72 mg PAC	100%	50%	Howell 2010	16 (Japan, Hungary)

Source: Howell AB, Botto H, Combescure C, et al. BMC Infectious Diseases 2010;10:94. PMID 20398248. Values are pooled medians from the mannose-resistant haemagglutination assay.

The statistics behind those medians are worth stating precisely. At 6 hours, 18 mg differed significantly from both 36 mg and 72 mg (p = 0.002), but 36 mg and 72 mg did not differ from each other. At 24 hours the picture inverted: the 72 mg arm still showed 50% anti-adhesion activity, against 12% for 36 mg and 0% for 18 mg (p = 0.002). The authors concluded that cranberry powder standardised to 72 mg of PAC per day “may offer some protection against bacterial adhesion and virulence in the urinary tract”, and described the effect as potentially nyctohemeral — covering the full day and night.

Note what this does not establish. Nobody in this trial took 36 mg twice a day. The 72 mg arm swallowed 72 mg in one go at breakfast. Splitting the dose is a sensible-sounding way to chase the same 24-hour coverage, and it is what several practitioners suggest, but it is an extrapolation from a single-dose experiment rather than a tested regimen. We would rather say that plainly than let a comparison table imply an evidence base that is not there.

Why “cranberry extract 500 mg” tells you nothing

Here is the practical problem. The number on the front of most cranberry products is an extract weight, and an extract weight has no fixed relationship to PAC content. The same arithmetic applied to drinks is in our piece on cranberry supplement versus juice. Two capsules both honestly labelled “cranberry extract 500 mg” can differ by an order of magnitude in the compound the research is actually about, for three separate reasons.

The extraction ratio. A 50:1 extract and a 4:1 extract are both extracts. They concentrate the source fruit to very different degrees, and unless the label states a drug-to-extract ratio you cannot tell which you are holding.

Standardisation, or its absence. A standardised product is manufactured to hit a stated percentage of a named marker compound. An unstandardised one is not. If no percentage and no marker appear on the panel, the PAC content is whatever that batch of fruit happened to contain.

The assay. PAC content is a measured value, and the measurement method changes the number. In the Howell trial, the PACs were quantified by Brunswick Laboratories using an updated dimethylaminocinnamaldehyde (DMAC) colorimetric method, run after open-column gel chromatography on Sephadex LH-20. That is a specific, documented procedure. A milligram figure produced by a different method is not automatically interchangeable with it. We looked for a published inter-assay variance figure we could quote here and could not verify one to our own standard, so we are not going to invent a percentage — the honest version is simply that the method matters and almost no consumer label names one.

There is a fourth wrinkle, less widely known: not all proanthocyanidins in a cranberry product are equally available. A 2021 paper in the Journal of Dietary Supplements compared urinary bacterial anti-adhesion activity after cranberry supplements containing soluble versus insoluble proanthocyanidins, and reported that the two forms did not behave the same way. We have not been able to verify that paper's author list and effect sizes against PubMed to the standard we apply to numbers on this site, so we cite it as a direction of travel rather than quoting figures from it: a milligram of “cranberry” is not automatically a milligram of usable PAC.

What your label does not tell you

We apply this to the product this site covers, because it would be dishonest not to. FemiCore does not publish a PAC milligram figure anywhere. We read this site's own ingredient breakdown and the manufacturer's sales-page text directly. Nine ingredients are named — cranberry extract, bearberry leaf, berberine, mimosa pudica seed and five Lactobacillus species — and not one milligram amount, standardisation percentage or CFU count appears for any of them. The manufacturer's own FAQ describes it as a proprietary formula.

That is not an accusation of anything; plenty of multi-ingredient supplements are sold this way. But it has a concrete consequence for the question this page is about: the 36 mg threshold cannot be checked against FemiCore, or against any other product that prints only ingredient names. A buyer who wants to know whether they are above or below the line that the research draws simply cannot find out.

If you want the full ingredient-by-ingredient picture including where we think the evidence is weakest, that is on our main FemiCore review. And if you are looking at a formula that also contains bearberry, read the duration limit regulators place on uva-ursi before you commit to a six-month supply — it is a separate issue from dose and it catches people out.

How to actually reach 36 mg

Four checks, in the order they are worth doing.

1. Find a PAC number, not an extract number. The panel should read something like “proanthocyanidins (PACs) 36 mg”. If it reads “cranberry fruit extract 500 mg” and stops, the product is unverifiable against the research, however good it may be.

2. Look for a named method. Better labels state how the PAC figure was measured. Most do not. A stated method is a signal that somebody measured rather than estimated.

3. Decide about 72 mg deliberately. Thirty-six milligrams is the threshold the pooled clinical data supports. Seventy-two milligrams is what carried measurable activity through to the next morning in the single dose-ranging trial. Those are two different claims resting on two different kinds of evidence, and it is worth knowing which one you are buying.

4. Plan for months, not weeks. The only duration band that reached significance in the 2024 pooled analysis was 12 to 24 weeks. A two-week trial of a cranberry product is not a test of anything.

What the evidence does not show

Being straight about the limits is the whole point of a dose article, so here they are.

Anti-adhesion activity is a surrogate, not an outcome. The Howell table above measures how well bacteria stick in a laboratory assay using a volunteer's urine. It does not count infections. A dose that wins on a surrogate marker has not thereby been shown to change how anyone feels.

The dose-ranging trial was small and short. Thirty-two healthy volunteers, a single morning dose, two urine collections. It was designed to compare doses, and it does that well. It was never designed to establish clinical benefit, and it did not.

The meta-analysis carries its authors' own caveats. Xiong and colleagues explicitly flag the limited quantity and quality of the available randomised trials, participant dropout that reduced statistical power, inconsistent criteria for what counted as an infection across the included studies, and follow-up capped at one year. An 18% relative reduction pooled from that evidence base is a real signal, not a settled fact.

The effect is not universal. The mixed-gender subgroup showed nothing. Neither did the sub-36 mg trials, whatever else was in them.

No dose is a treatment. None of this concerns an active infection. Cranberry proanthocyanidins are studied as a preventive support measure in people prone to repeat episodes. If you have symptoms — burning, urgency, fever, pelvic or flank pain, or blood in the urine — that is a medical situation and a clinician is the answer, not a capsule.

And no finished product has been trialled. Ingredient evidence is not product evidence. There is no published clinical trial of FemiCore itself, and there is no PAC figure on its label to connect it to the trials of the ingredient. The same gap runs through most of the category, which is why our overview of urinary support supplements for women ranks ingredients rather than brands.

Frequently asked questions