MERIT: D-Mannose versus Placebo for Prevention of Recurrent UTI in Primary Care
MERIT was a multicenter, double-blind, placebo-controlled randomized trial in 598 women with recurrent urinary tract infection (UTI) from 99 UK primary care centers that compared D-mannose 2 g daily with a fructose placebo for 6 months. A further clinically suspected UTI with an ambulatory care contact occurred in 150 of 294 women (51.0%) with D-mannose and 161 of 289 (55.7%) with placebo (risk difference −5 percentage points, 95% CI −13 to 3; P = 0.26).[1]
- Design and population
- Multicenter, double-blind, placebo-controlled superiority randomized trial. 598 women aged 18 or older (mean age 58) with at least 2 UTIs in the preceding 6 months or 3 in 12 months, recruited from 99 primary care centers in England and Wales.
- Arms
- D-mannose powder 2 g daily versus fructose placebo powder
- Report and follow-up
- 2024 · 6 months per participant; recruitment March 28, 2019 to January 31, 2020
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Proportion with at least 1 further episode of clinically suspected UTI for which ambulatory care was contacted within 6 months; unadjusted risk difference, D-mannose minus placebo, in percentage points | 150/294 (51.0%) vs 161/289 (55.7%) | difference -5; CI -13–3; P 0.26 | Not met |
| Number of clinically suspected UTIs over 6 months; adjusted incidence rate ratio, D-mannose versus placebo | mean 0.9 (SD 1.19) vs mean 1.0 (SD 1.28) | RR 0.88; CI 0.72–1.08; P 0.21 | Not met |
Harms: Serious adverse events (events, excluding hospitalizations for elective procedures): 20 vs 8; Adverse effects given as the reason for discontinuing the study product: 10/303 vs 5/295
ISRCTN13283516 · Report DOI
Design
Women were eligible if they were 18 or older and had at least 2 UTIs in the preceding 6 months or 3 in the preceding 12 months, documented as consultations with UTI symptoms or a UTI-specific antibiotic prescription in the primary care record. Exclusions included pregnancy or planned pregnancy, a diagnosis of interstitial cystitis or overactive bladder syndrome, nursing home residence, any catheterization including intermittent self-catheterization, and use of the immunostimulant Uromune. Women who had started antibiotic prophylaxis in the past 3 months and would not stop it, or who were already taking D-mannose and would not stop, were also excluded. Of 7591 women invited, 830 responded and 598 were randomized in equal allocation, with block randomization stratified by practice; participants, clinicians and the team responsible for recruitment and follow-up were masked. Recruitment ran from March 28, 2019 to January 31, 2020.[1]
The primary outcome was the proportion of women with at least 1 further episode of clinically suspected UTI for which they contacted ambulatory care (primary care, emergency department, hospital, ambulance or out-of-hours services) within 6 months, taken from the primary care record. The outcome was chosen with women who had lived experience of recurrent UTI and does not require microbiological confirmation. The sample size assumed a 26.6% UTI rate in the control group from a 2016 cranberry trial and aimed to detect a 50% relative reduction, the smallest effect the patient advisory panel would accept for committing to daily prophylaxis, with 90% power and a two-sided alpha of 0.05. That required 508 women with allowance for 20% loss to follow-up; the target was raised to 598 on November 7, 2019 because interim blinded data showed more withdrawals than anticipated. The primary analysis included all randomized women with primary outcome data, by allocation.[1]
Mean age was 58.0 years, 376 women (63.1% of those answering) reported having gone through the menopause, and the median number of self-reported UTIs in the previous 12 months was 4.0 (interquartile range 3.0 to 6.0). At enrollment 142 women (23.9% of those answering) reported current UTI symptoms.[1]
Results
Primary outcome data were available for 294 of 303 women randomized to D-mannose (97.0%) and 289 of 295 randomized to placebo (98.0%). A further clinically suspected UTI with an ambulatory care contact occurred in 150 (51.0%) and 161 (55.7%), a risk difference of −5 percentage points (95% CI −13 to 3; P = 0.26) and an adjusted relative risk of 0.92 (95% CI 0.80 to 1.05; P = 0.22). The estimate was similar in per-protocol analyses, in analyses imputing missing outcomes as positive or negative, and in preplanned subgroups by UTI frequency and menopausal status. The authors note that the 95% confidence intervals exclude the minimum effect the patient panel considered important.[1]
The number of clinically suspected UTIs, the key secondary outcome, did not differ (adjusted incidence rate ratio 0.88, 95% CI 0.72 to 1.08; mean 0.9 versus 1.0 per woman). Time to the next consultation for a clinically suspected UTI gave an adjusted hazard ratio of 0.86 (95% CI 0.69 to 1.08), with 4.11 versus 4.80 episodes per 1000 days at risk. Microbiologically proven UTIs (incidence rate ratio 0.97, 95% CI 0.63 to 1.48) and antibiotic courses prescribed for UTI (0.88, 95% CI 0.69 to 1.12) were also similar. Days of moderately bad or worse symptoms in the diary had an adjusted median difference of 0.00 (95% CI −0.36 to 0.36).[1]
Days of prescribed antibiotics were lower with D-mannose (adjusted median difference −3.00 days, 95% CI −4.40 to −1.60). The report attributes this to a few women with very high consumption (up to 161 days in the D-mannose group and 66 days in the placebo group); in a post hoc analysis that excluded 12 women more than 3 SDs from the mean, the means were 4.7 and 5.1 days with an adjusted median difference of 0.00 (95% CI −1.40 to 1.40). Resistant uropathogens were cultured during infection in 41 of 294 women (13.9%) with D-mannose and 45 of 289 (15.6%) with placebo (relative risk 0.90, 95% CI 0.60 to 1.33), and UTI-related hospital admissions occurred in 6 (2.0%) and 4 (1.4%).[1]
Surgery and treatment delivery
Women took a daily scoop of about 2 g of D-mannose powder or a similar scoop of fructose powder; both were white and sweet, supplied in identical bottles, and sent in packs every 2 months. They continued the product during symptomatic episodes and while taking antibiotics. Of those randomized, 294 of 303 (97.0%) and 283 of 295 (95.9%) started the product. More than two-thirds of women reported taking it at least 3 days a week for 15 weeks. At 6 months, 75 of 224 women (33.5%) in the D-mannose group and 64 of 215 (29.8%) in the placebo group believed they had received D-mannose (odds ratio 1.19, 95% CI 0.79 to 1.78).[1]
Adverse events
Twenty serious adverse events were reported in the D-mannose group and 8 in the placebo group, 28 in total among the 598 randomized women; none was judged related to the intervention. Adverse effects were given as the reason for stopping the product in 10 women in the D-mannose group and 5 in the placebo group in the flow diagram, and an adverse event that prevented compliance with trial procedures in 1 and 2.[1]
Limitations
The dose was delivered as a powder, which the authors note may have produced less consistent daily dosing than capsules, and adherence was self-reported. The primary outcome was a clinical diagnosis recorded after an ambulatory care contact, not a culture-confirmed infection, although urine cultures were sent at more than 60% of contacts for suspected UTI and the proportion of women with at least 1 microbiologically confirmed UTI was similar in the two groups. Antibiotic-use outcomes were skewed by a small number of heavy users. The trial was not powered for subgroups, including premenopausal women, and the authors state that the results do not necessarily apply to secondary care populations who may be more severely affected. The Results text and the flow diagram give different numbers of withdrawn or discontinued women (69 and 65 versus 64 and 64).[1]
In conversation
Same setting ALTAR (2022)
The earlier D-mannose trials had no placebo comparison and enrolled women at the time of a symptomatic infection. As summarized in the MERIT report, a 308-woman three-arm trial found 6-month recurrence of 14.6% with D-mannose 2 g daily, 20.4% with nitrofurantoin 50 mg daily and 60.8% with no prophylaxis[4], and a 60-woman crossover trial found a mean time to recurrence of 200 days with D-mannose versus 52.7 days with co-trimoxazole.[1] MERIT randomized women with a history of recurrent UTI, tested D-mannose against placebo and recorded outcomes from the primary care record.[1] ALTAR tested another non-antibiotic agent, methenamine hippurate, against daily antibiotic prophylaxis in an open-label noninferiority design in secondary care, so the two trials answer different questions about alternatives to antibiotics.
Guidelines and approvals
- AUA/CUA/SUFU (Recurrent Uncomplicated UTI in Women, 2019; confirmed 2022; amended 2025), statement 14. Clinicians should inform patients with recurrent UTI that D-mannose alone for prophylaxis may not be effective in UTI prevention . The rationale cites MERIT as a high-quality large randomized trial (N = 598) with no difference in UTI recurrence between D-mannose 2 g per day and placebo (relative risk 0.92, 95% CI 0.80 to 1.05) and says there is little evidence to support its benefit in preventing UTI.[2]
- EAU (Urological Infections, limited update March 2026), section 3.5.4. Use D-mannose to reduce recurrent cystitis episodes, but patients should be informed of the overall weak and contradictory evidence of its effectiveness . The guideline text describes MERIT and its finding that daily D-mannose did not reduce the proportion of women with recurrent cystitis in primary care who had a subsequent clinically suspected cystitis.[3]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2024-04-08 | Primary report (JAMA Intern Med) · primary | — | Proportion with at least 1 further episode of clinically suspected UTI for which ambulatory care was contacted within 6 months; unadjusted risk difference, D-mannose minus placebo, in percentage points: 150/294 (51.0%) vs 161/289 (55.7%); difference -5; CI -13–3; P 0.26 Not met Number of clinically suspected UTIs over 6 months; adjusted incidence rate ratio, D-mannose versus placebo: mean 0.9 (SD 1.19) vs mean 1.0 (SD 1.28); RR 0.88; CI 0.72–1.08; P 0.21 Not met Adjusted hazard ratio, D-mannose versus placebo (mixed-effects Cox model); incidence per 1000 days at risk: 4.11 per 1000 days at risk vs 4.80 per 1000 days at risk; HR 0.86; CI 0.69–1.08; P 0.20 Symptom-diary days over 6 months; adjusted median difference, D-mannose minus placebo (quantile regression): median 0.0 (IQR 0.0 to 2.0); mean 1.7 vs median 0.0 (IQR 0.0 to 2.0); mean 3.2; difference 0; CI -0.36–0.36; P >0.99 Number of microbiologically proven UTIs over 6 months; adjusted incidence rate ratio, D-mannose versus placebo: mean 0.4 (SD 0.87) vs mean 0.4 (SD 0.78); RR 0.97; CI 0.63–1.48; P 0.88 Number of prescribed antibiotic courses for UTI over 6 months; adjusted incidence rate ratio, D-mannose versus placebo: mean 0.9 (SD 1.32) vs mean 1.1 (SD 1.53); RR 0.88; CI 0.69–1.12; P 0.29 Days of antibiotics prescribed for UTI over 6 months; adjusted median difference, D-mannose minus placebo (quantile regression): median 0.0 (IQR 0.0 to 7.0); mean 5.7 (SD 12.49) vs median 3.0 (IQR 0.0 to 7.0); mean 6.3 (SD 10.28); difference -3; CI -4.4–-1.6; P <0.001 Post hoc analysis excluding 12 women more than 3 SDs from the mean; mean days (SD); adjusted median difference, D-mannose minus placebo: 4.7 (SD 7.1) vs 5.1 (SD 7.5); difference 0; CI -1.4–1.4; P >0.99 Women with resistant uropathogens cultured during an episode of infection; relative risk, D-mannose versus placebo: 41/294 (13.9%) vs 45/289 (15.6%); RR 0.9; CI 0.6–1.33; P 0.59 Women admitted to hospital for UTI during 6 months; relative risk, D-mannose versus placebo: 6/294 (2.0%) vs 4/289 (1.4%); RR 1.47; CI 0.47–4.61; P 0.51 | Hayward G, Mort S, Hay AD, et al. "d-Mannose for Prevention of Recurrent Urinary Tract Infection Among Women: A Randomized Clinical Trial." *JAMA Intern Med.* 2024;184(6):619-628. doi:10.1001/jamainternmed.2024.0264 DOI |
References
1. Hayward G, Mort S, Hay AD, et al. "d-Mannose for Prevention of Recurrent Urinary Tract Infection Among Women: A Randomized Clinical Trial." JAMA Intern Med. 2024;184(6):619-628. doi:10.1001/jamainternmed.2024.0264
2. American Urological Association, Canadian Urological Association and Society of Urodynamics, Female Pelvic Medicine & Urogenital Reconstruction. Recurrent Uncomplicated Urinary Tract Infections in Women: AUA/CUA/SUFU Guideline. 2019; confirmed 2022; amended 2025. Statement 14.
3. Bonkat G, Kranz J, Cai T, et al. EAU Guidelines on Urological Infections. Arnhem: EAU Guidelines Office; limited update March 2026. Sections 3.5.3.c.7 and 3.5.4.
4. Kranjčec B, Papeš D, Altarac S. "D-mannose powder for prophylaxis of recurrent urinary tract infections in women: a randomized clinical trial." World J Urol. 2014;32(1):79-84. doi:10.1007/s00345-013-1091-6