D-Mannose vs Cranberry PACs: Which Actually Has the Evidence in 2026?

The head-to-head evidence table

This is the comparison almost nobody publishes. Most pages give you two prose sections and leave you to hold them side by side in your head. Here are both compounds against the same five columns, using each one's largest and most rigorous trial evidence rather than its most flattering.

Compound	Dose used in the pivotal trial	Trial + year	n	Primary outcome result	2025 guideline position
D-mannose	2 g/day powder, 6 months	MERIT, Hayward et al., JAMA Intern Med 2024	598 women (mean age 58), 99 UK primary-care sites	51.0% (150/294) contacted care for a suspected episode vs 55.7% (161/289) on placebo. Risk difference −5% (95% CI −13% to +3%), P = 0.26 — not significant.	AUA/CUA/SUFU 2025: clinicians should inform patients that d-mannose alone may not be effective for prevention.
Cranberry PACs	Varies by trial; benefit concentrated at ≥36 mg PAC/day	Williams et al., Cochrane Database Syst Rev 2023, CD001321.pub7	50 trials, 8,857 randomised participants	Overall RR 0.70 (95% CI 0.58–0.84). Women with recurrent episodes: RR 0.74 (95% CI 0.55–0.99), 8 studies, 1,555 participants, moderate certainty.	AUA/CUA/SUFU 2025: recommendation strengthened from may offer to should offer for prophylaxis.
Cranberry PACs, dose split	≥36 mg/day vs <36 mg/day	Xiong et al., Front Nutr 2024;11:1422121	10 RCTs, 2,438 participants	≥36 mg/day: RR 0.82 (95% CI 0.69–0.98), P = 0.03. Below 36 mg/day: no statistically significant reduction (P = 0.39).	Not a guideline document — this is the dose evidence sitting underneath the guideline.

What the MERIT trial actually did

MERIT is the trial that changed the d-mannose conversation, and it is the trial that the pages currently ranking for this comparison do not cite. It was published in JAMA Internal Medicine in 2024 by Hayward and colleagues, and it is the largest placebo-controlled randomised trial of d-mannose ever run.

The design was deliberately pragmatic. Researchers recruited 598 women through 99 UK primary-care practices — ordinary general practices, not a specialist clinic with an unusually motivated population. The women had a documented history of repeat episodes: at least two in six months or three in twelve. Mean age was 58, with a range from 18 to 93. Half received 2 g of d-mannose powder daily; half received a matched placebo. Everyone was followed for six months. Neither the women nor the researchers knew who was taking what.

The result was flat. In the d-mannose group, 150 of 294 women (51.0%) contacted care for a suspected urinary infection during the six months. In the placebo group, 161 of 289 (55.7%) did. The risk difference was −5%, with a 95% confidence interval running from −13% to +3%, and a P value of 0.26. That confidence interval crossing zero is the whole story: the data are entirely compatible with d-mannose doing nothing.

The authors' own conclusion was blunt — d-mannose should not be recommended for prophylaxis in this group. That is a strong statement from a trial team, and it deserves to be reported accurately rather than buried.

Why the earlier d-mannose studies looked better

Smaller studies from the 2010s produced far more optimistic numbers, and they are the ones that built d-mannose's reputation. The honest reading is not that those studies were fraudulent, but that they were smaller, several were open-label rather than blinded, and some compared d-mannose against an antibiotic rather than against a placebo — a design that can make an ineffective agent look acceptable if the comparator also underperforms. When a large, blinded, placebo-controlled trial arrives and finds nothing, it generally outweighs a stack of smaller and weaker ones. That is how the evidence hierarchy is supposed to work, and it is uncomfortable when the answer goes against a popular supplement.

What the 2023 Cochrane review found — including where cranberry failed

The Cochrane review of cranberry for preventing urinary tract infections was updated in November 2023 (Williams et al., CD001321.pub7). It pooled 50 randomised or quasi-randomised trials covering 8,857 participants, which makes it the single largest body of evidence on either compound discussed here.

The headline number is an overall risk ratio of 0.70 (95% CI 0.58–0.84). For the group most relevant to a woman reading this page — women with a history of recurrent episodes — the risk ratio was 0.74 (95% CI 0.55–0.99), drawn from 8 studies and 1,555 participants, rated moderate certainty. A risk ratio of 0.74 is a 26% relative reduction. That is the figure that circulates on supplement blogs, and it is real, but it belongs to that subgroup and only that subgroup.

What almost nobody quotes is the other half of the same review. Cranberry did not work everywhere:

Population	Risk ratio (95% CI)	Studies / participants	Cochrane's reading
Women with recurrent episodes	0.74 (0.55–0.99)	8 / 1,555	Probably reduced risk
Children	0.46 (0.32–0.68)	5 / 504	Reduced risk
People susceptible after an intervention	0.47 (0.37–0.61)	6 / 1,434	Reduced risk
Elderly people in institutions	0.93 (0.67–1.30)	3 / 1,489	Little or no benefit
Pregnant women	1.06 (0.75–1.50)	3 / 765	Little or no benefit
Adults with bladder emptying problems	0.97 (0.78–1.19)	3 / 464	Little or no benefit

Three of the six subgroups show nothing. Cranberry is not a general-purpose urinary supplement that helps everyone who takes it; it is an intervention with a defined population where the data support it and several populations where they do not. Any page that quotes the 26% figure without that table is selling, not reporting.

Cochrane also compared cranberry products against antibiotics directly and found little or no difference between them (RR 1.03, 2 studies, 385 participants) — a small comparison, and one that should be read as a signal for further research rather than as a reason to substitute one for the other. Nobody should stop a prescribed medicine on the strength of two small trials, and the 2025 recurrent-UTI guideline does not suggest otherwise. The other free lever with randomised evidence behind it is fluid volume, covered in water intake and urinary health.

What the 2025 guideline amendment changed

In 2025 the American Urological Association, together with the Canadian Urological Association and the Society of Urodynamics, Female Pelvic Medicine & Urogenital Reconstruction, published an amendment to their guideline on recurrent uncomplicated urinary tract infections in women. Two changes are directly relevant here.

First, the cranberry recommendation was strengthened. The previous language said clinicians may offer cranberry prophylaxis; the amendment moves it to clinicians should offer it, on the basis of newer and stronger evidence. That is a meaningful shift in guideline vocabulary — should and may are not interchangeable in these documents.

Second, the guideline addresses d-mannose explicitly, advising clinicians to inform patients that supplements containing only d-mannose may not be effective for prevention. The panel had MERIT available; the pages currently ranking for this comparison were written before it.

Two important framing notes. This is reporting on what a clinical guideline says, not a claim about any product. A guideline recommendation about cranberry as an ingredient class is not a statement about any particular capsule, including FemiCore, and no dietary supplement prevents, treats or cures any disease. Read the guideline as context for a conversation with your own clinician, which is exactly what it is written for.

The dose question everyone skips

Here is the part that makes most of this comparison academic: cranberry's evidence is dose-dependent, and most cranberry labels do not print the dose that matters.

A 2024 meta-analysis in Frontiers in Nutrition (Xiong et al., 11:1422121) pooled 10 randomised trials covering 2,438 participants and split them by proanthocyanidin intake. At 36 mg PAC per day or more, risk fell by 18% (RR 0.82, 95% CI 0.69–0.98, P = 0.03). Below 36 mg per day, there was no statistically significant reduction at all (P = 0.39). In female-only cohorts the reduction was RR 0.84 (95% CI 0.71–0.98, P = 0.02). The same analysis found the effect only reached significance when the product was used for 12 to 24 weeks — which tells you something about how long a fair trial of any of this takes.

That is a threshold, not a gradient. It means a cranberry product below the line is not a weaker version of one above it; in the pooled data it is a product with no demonstrated effect. And a label that says cranberry extract 500 mg tells you nothing about which side of that line it sits on, because PAC content depends on the extraction and the standardisation, not on the total weight. Our review of the underlying cranberry and Lactobacillus evidence covers whether these ingredients work; this page is about which of the two candidate compounds the data actually back.

We should apply that standard to the product this site covers, and it does not pass. FemiCore names cranberry extract as an ingredient, but neither this site's review nor the official sales page publishes a proanthocyanidin milligram figure — or any milligram figure, for any of its nine ingredients. We checked both directly. That means nobody, including us, can tell you whether FemiCore's cranberry content clears the 36 mg threshold. It is an honest gap, and the same gap applies to a very large share of the cranberry products on the market. The same disclosure problem affects the probiotic side of these formulas, which we cover in our audit of what L. crispatus labels actually deliver.

What this evidence does not show

Five limitations, stated plainly, because the pages competing for this query mostly skip them.

It does not show that d-mannose is inert. MERIT tested one dose (2 g daily), one formulation (powder), one duration (six months) and one population (women with documented recurrent episodes, mean age 58). A null result under those conditions is strong evidence about those conditions. It is not proof that no dose, format or population would respond. It is, however, the best evidence we have, and it points one way.

It does not show that cranberry works for you specifically. A risk ratio of 0.74 is a population average. Some women in those trials had no episodes and some had several; the average moved modestly. Nothing in a pooled estimate predicts an individual outcome.

It does not show that any finished multi-ingredient product works. Ingredient evidence is not product evidence. FemiCore has no clinical trial on the finished formula, and we have said so consistently. A capsule containing an ingredient with trial data behind it inherits the plausibility of that data, not its results.

It does not show a combination benefit. No trial has tested cranberry plus d-mannose against either alone. Marketing that implies the pair is additive is asserting something no one has measured. For how these two sit against the rest of the shelf — probiotics, uva-ursi and plain water included — see our ranking of urinary support supplements for women.

It does not show anything about treatment. Every trial discussed here measured prevention in people prone to repeat episodes. None tested clearing an existing infection, and a dietary supplement is not a substitute for antibiotics when they are needed.

Frequently asked questions