Whole cranberry fruit powder supplement reduces the incidence of culture-confirmed urinary tract infections in females with a history of recurrent urinary tract infection: A 6-month multicenter, randomized, double-blind, placebo-controlled trial

Abstract

Background

High prevalence of urinary tract infections (UTI), including cystitis, and concern for antimicrobial resistance justify safe and effective nonantibiotic therapies for prevention of recurrent UTI (rUTI).

Objectives

This study investigated the effect of a whole cranberry fruit powder supplement on incidence of culture-confirmed UTI (primary outcome) in females with rUTI history.

Methods

This multicenter, 6-mo, randomized, placebo-controlled, double-blind study enrolled 150 healthy females [18–65 y, body mass index (BMI) >17.5 and <35 kg/m²] with rUTI defined as ≥3 UTIs in the last year or ≤2 UTIs in the last 6 mo, excluding those with >5 UTIs in the last 6 mo. Participants consumed either 1 capsule of 500 mg/d of whole cranberry powder (Pacran) or placebo. Culture-confirmed UTIs (>10⁸cfu/L) were assessed throughout the intervention period at unscheduled clinic visits whenever participants experienced UTI symptoms and at baseline, 3- and 6-mo clinic visits. Symptomatic suspected UTIs were defined as participant-reported UTI-associated symptoms at unscheduled visits.

Results

Whole cranberry powder capsules reduced culture-confirmed UTI risk compared with placebo by 52% (adjusted relative risk [RR]: 0.48; 95% confidence interval [CI]: 0.26, 0.87; P = 0.01); reduced Escherichia coli UTIs (RR: 0.49; 95% CI: 0.24, 1.01; P = 0.05); reduced incidence of UTI with urinary frequency and urgency symptomatology (RR: 0.29; 95% CI:0.13, 0.63; P < 0.01); delayed time to first UTI episode (adjusted hazard ratio [HR]: 0.36; 95% CI: 0.18, 0.74; P = 0.01); and reduced the mean total number of UTIs per participant (adjusted incidence rate ratio IRR: 0.41; 95% CI: 0.21, 0.79; P = 0.01). Significant differences between groups in incidence of symptomatic suspected UTIs and culture-confirmed dysuria were not observed. Exploratory scores for UTI-related female sexual matters, assessed in a subset of sexually active, consenting females, did not differ significantly between groups. No safety concerns were reported.

Conclusion

This study shows that whole cranberry powder capsules do not impact safety markers and reduce the incidence of culture-confirmed UTI and several other UTI-related outcomes in healthy females with rUTI history.

This trial was registered at clinicaltrials.gov asNCT03042273.

Introduction

Globally, urinary tract infections (UTIs) are among the most common bacterial infections in clinical practice, resulting in substantial clinical and economic burdens [[1], [2], [3]]. Females are more susceptible, with nearly 1 in 3 females having at least one episode of UTI requiring antimicrobial therapy by the age of 24 y, ∼50 to 70% experiencing a UTI during their lifetime, and ∼20 to 40% of those who have had a UTI will have recurrent UTI (rUTI) [1,[4], [5], [6]]. UTIs are associated with pain and discomfort, which can cause stress and affect a patient’s quality of life [2,7]. The most common form of UTI is cystitis, an acute uncomplicated infection confined to the bladder. Cystitis remains one of the most common conditions requiring antimicrobial treatment, which is also highly effective as a prophylactic measure [[8], [9], [10]]. However, continued use is often associated with antimicrobial resistance [11]. The impact on antimicrobial resistance rates and subsequent health care and economic burden is of major concern [2,12] such that establishing alternative safe and effective nonantibiotic therapies for prevention of rUTI in otherwise healthy females should be prioritized. Cranberry (Vaccinium macrocarpon) products have been suggested as an alternative therapeutic strategy. Components in the cranberry fruit, including proanthocyanidins and flavonoids, have been shown in both in vitro and ex vivo experiments to induce antiadhesive activity against uropathogenic Escherichia coli to uroepithelial cells [[13], [14], [15], [16]], resulting in impaired colonization and reduced subsequent infection [17]. The most recent Cochrane review and meta-analysis of randomized controlled trials that evaluated the effects of various cranberry-containing products on risk of UTIs in susceptible populations reported reduced risk of symptomatic, culture-confirmed UTIs in females with rUTIs with moderate certainty (8 studies, 1555 participants: relative risk [RR]: 0.74; 95% confidence interval [CI]: 0.55, 0.99; I² = 54%) [18]. While, most clinical trials to date have tested fractions from cranberry fruit, such as juice concentrate or isolated proanthocyanidins [19], two clinical studies [20,21] investigated a powder consisting of whole cranberry fruit. Although they both described reduced incidence of UTI in females with rUTI, the trials presented methodological and statistical limitations [22]. The study by Sengupta et al. [20] was short (90 d), small (n = 60 across 3 groups), treatments were not blinded, the primary outcome (presence of urinary Escherichia coli) was only assessed at visits and not as infections occurred, and unequal positive cultures were observed between groups at baseline. The study by Vostalova et al. [21], although generally well designed, was criticized by the European Food Safety Authority for not including a proper intention-to-treat analysis [22]. Therefore, high quality randomized controlled trials to assess the efficacy and safety of whole cranberry fruit powder to reduce incidence of UTIs in females with rUTIs are warranted.

The primary aim of the current study was to investigate the effectiveness of a whole cranberry fruit powder on the incidence of culture-confirmed UTI in females with a history of rUTI (≥3 UTIs in the last year or ≤2 UTIs in the last 6 mo, but not >5 UTIs), over a 6-mo intervention period. Secondary outcomes included incidence of symptomatic suspected UTI, incidence of individual symptoms of culture-confirmed UTI (dysuria, urinary frequency or urgency, fever, suprapubic pain, and macroscopic hematuria), time to first culture-confirmed UTI episode, number of culture-confirmed UTIs per participant and presence of individual uropathogenic species on culture at any time of a culture-confirmed UTI and safety assessments. Sexual health was investigated as an exploratory outcome in a subset of sexually active and consenting females.

Methods

The trial was conducted at 5 sites across Australia, including the Commonwealth Scientific Industrial Research Organization (CSIRO) Nutrition and Health Research Clinic, Adelaide, South Australia; Holdsworth House Medical Centre, Sydney, New South Wales and Brisbane, Queensland; Griffith University Clinical Trial Unit, Gold Coast, Queensland; and Monash Alfred Psychiatry Research Centre (MAPrc), Melbourne, Victoria. Human Research Ethics Committee approvals were obtained from the CSIRO Human Research Ethics Committee (Adelaide, South Australia) (reference number: 23_2016) and Bellberry Limited (Adelaide, South Australia, Australia) (reference number: 2017-12-933). Oral and written informed consent were obtained. The trial was prospectively registered with clinicaltrials.gov (NCT03042273). The intervention phase was executed from 15 May 2017 to 31 December 2019. Important changes to the methods after trial commencement included introducing a BMI inclusion criterion to ensure the recruitment of healthy participants and prematurely ending recruitment after enrolling 150 of the planned 300 participants due to slow recruitment rates and the determination that the target could not be achieved within an extended timeframe. Additional methodological details are provided in the Supplementary materials (Supplementary Methods).

Participants

Females with a history of rUTI were recruited via clinic site databases, flyers at shopping center washrooms, local hospitals, pharmacies, medical practices/urologist clinics and other local venues, radio advertisements, newspaper articles and advertisements, social media, and recruitment agencies. Inclusion criteria were females, aged 18–65 years; history of rUTI (e.g., recurrent cystitis) defined as ≥3 UTIs in the last year or ≤2 UTIs in the last 6 mo; and BMI >17.5 and <35 kg/m². To ensure the recruitment of participants with rUTI rather than other urologic conditions, the following evidence was required to confirm rUTI: a case history from a healthcare professional, including dates of infections, prescribed medical treatments, and ≤1 positive urinary culture pathology report over the last 12 mo. In the absence of urinary culture reports, eligibility was determined by the medical investigator based on the provided evidence. Exclusion criteria were microbial growth on urine culture [≥10⁷ cfu/L (10⁴ cfu/mL)] 7 d prebaseline to ensure that participants were free of UTIs at the start of the study; history of >5 UTIs in last 6 mo as females experiencing higher number of UTIs may not be responsive to improvement through dietary supplementation [23]; the use of antibiotics, including prophylactic antibiotics for treating a UTI, or any antibacterial products that could interfere with the study outcomes, within 28 d prior to baseline; regular use of Vaccinium containing products (e.g., all forms of blueberries, cranberries, bilberry, lingonberry, etc. including fruit, dried fruit, pills, juices, or supplements) 28 d prebaseline; intermittent or indwelling urinary catheter; anatomical abnormalities of urinary tract; history of or known clinically significant renal or urologic disease; pregnant and/or breastfeeding and not using effective methods of contraception; history of or known clinically significant cardiac, liver, gastrointestinal, metabolic disease or diabetes; immunocompromised or taking warfarin; history of or known presence of alcohol abuse or illicit drug use; current or planned hospitalizations; and history of adverse reaction, hypersensitivity or suspected allergy to investigational product ingredients.

Study design and procedures

This trial was designed as a 6-mo multicenter, placebo-controlled, double-blind, parallel-arm phase-II study. Eligible participants were randomly assigned to receive whole cranberry or placebo in a 1:1 ratio, stratified by 5 study sites and age groups (18–45 y; >45 y) using randomly permuted blocks of size 2 and 4. The randomized allocation schedule was provided by the National Health and Medical Research Council Clinical Trials Centre Central Randomization Service, Camperdown, New South Wales, Australia. Treatments were labeled with sequential kit numbers according to the randomized allocation schedule. Once a participant’s eligibility was confirmed, the investigational site contacted the National Health and Medical Research Council Clinical Trials Center Central Randomization Service via an interactive voice response system to obtain the participant’s unique randomization identifier and treatment kit number. Whole cranberry powder and placebo capsules were matched for appearance. Treatment allocation was further concealed by sealing the capsules in identical opaque containers numbered sequentially. Participants, all study staff, and statisticians were blinded to treatment. The schedule of assessments is outlined in Supplemental Table 1.

Culture-confirmed UTIs (>10⁸cfu/L) were assessed throughout the intervention when participants experienced UTI symptoms and at baseline, 3- and 6-mo clinic visits. Participants who experienced UTI symptoms, including dysuria, urinary frequency or urgency, fever, suprapubic pain, and macroscopic hematuria, attended an in-clinic unscheduled visit for urinalysis and culture and were prescribed antibiotic treatment in line with standard medical practice. Participants unable to attend the clinic for an unscheduled visit were requested to drop their urine sample off at their local pathology laboratory using presupplied “emergency kits” (containing an antibiotic prescription, thermometer, urinary collection sterile pot, and urine pathology form) and then fill the prescription. They then attended an inclinic unscheduled visit on the next business day. Online surveys were conducted at 1-, 2-, 4-, and 5-mo to assess treatment compliance, adverse events (AEs), and use of concomitant medications. A final participant online survey was conducted 28 d after the 6-mo study visit for a final safety assessment, including AE and concomitant medications. Compliance with study treatments was assessed by having participants bring their capsule containers to their 3- and 6-mo clinic visits. Compliance was calculated as the percentage of capsules consumed over the 6-mo period relative to the number of capsules that should have been consumed.

Investigational products

The commercial cranberry supplement (Swisse High Strength Cranberry [Pacran, Givaudan Flavors Corp.]) contained 500 mg whole cranberry powder Pacran and a standardized concentration of proanthocyanidins >1.5% by HPLC (>7.2% via European Pharmacopoeia and >0.3% via Brunswick Laboratories 4-dimethylaminocinnamaldehyde assay). The placebo contained color-matched soy oil. Other ingredients in both whole cranberry powder and placebo capsules included soy oil, lecithin, hydrogenated vegetable oil, and beeswax. Participants consumed 1 capsule daily during or immediately after a main meal and ideally at the same time every day for 6 mo. Active and placebo capsules were identical in size, shape, color, and smell.

Urinalysis and urine culture

A midstream urine sample was collected for dipstick, urinalysis, and culture throughout the intervention period whenever participants experienced UTI symptoms and at baseline, 3- and 6-mo clinic visits. Culture and urine pathology analyses were performed by National Association of Testing Authorities, Australia laboratories at each study site. Uropathogenic species were determined within the urine sample and included: Escherichia coli, Enterococci, Enterobacter spp., Staphylococcus saprophyticus, Klebsiella pneumoniae, Coagulase negative staphylococci, Enterococcus faecalis, and Klebsiella oxytoca. Incidence of culture-positive UTI (primary outcome) was assessed by (1) culture confirmation at a level of >10⁸ cfu/L (10⁵ cfu/mL) in participants who experienced UTI symptoms and attended an unscheduled visit and (2) blind panel review of culture-confirmed or mixed urine cultures picked up at routine study visits. The blind review panel determined whether they were culture-confirmed UTIs using the presence of symptoms in combination with urinalysis data (illustrated in Supplemental Figure 1). Where a clear decision could not be made, the results were taken as positive based on cultural result. Symptomatic suspected UTI was defined as attending an unscheduled visit with participant-reported UTI-associated symptoms, with or without culture-positive UTI.

Assessments of safety parameters

The following safety parameters were assessed (methods are described in Supplementary material): vital signs (blood pressure, pulse rate, respiratory rate, and body temperature) and physical examination at baseline, 6-mo clinic and unscheduled visits; fasting blood hematology and biochemistry parameters at baseline, 3- and 6-mo clinic visits; incidence of AE and serious adverse events (SAEs) and its severity and causality to study treatment and, use of concomitant medications at baseline, 3- and 6-mo clinic, unscheduled visits and online surveys at 1-, 2-, 4-, and 5-mo.

Statistical analyses

Data collected during this study were processed and managed as per a prespecified data management plan. The original sample size calculation indicated n = 300 females (150 participants in each group) provided ≥80% power to detect an absolute reduction of 15% in the rate of culture-confirmed UTI incidence between treatment groups (2-sided alpha = 0.05), given expected control group incidence is 25%–35% and dropout rate of ≤10%. At the time when recruitment ceased, n = 150 patients had been randomly assigned (75 participants in each treatment group). It was calculated that a revised sample size of n = 150 (75 participants in each group) provided 80% power to detect an absolute reduction of ∼19% in the rate of culture-confirmed UTI incidence between treatment groups (2-sided alpha = 0.05) given an incidence of 30%. No interim analysis was conducted.

Primary efficacy analyses were conducted according to the intention-to-treat principle on the full analysis set (FAS) comprising all randomly assigned participants with ≥1 dose of study treatment and with ≥1 nonmissing postbaseline value for the primary outcome [24] and according to their randomly assigned treatment allocation. Following recommendations of Jakobsen et al. [25], missing outcome data were not imputed, given the proportions of missing outcome data were small (<5%) and baseline covariate data were complete. The FAS, thus, included all randomly assigned individuals with both baseline data and valid outcome data and who took ≥1 dose of study medication. Secondary per-protocol analyses were performed on the per-protocol set comprising all randomly assigned participants who were compliant with treatment and reported no major deviations or violations with respect to the protocol (see Supplementary material for assessment of protocol compliance and major protocol deviations or violations), according to the study treatment received. Analyses of safety outcomes were conducted on the safety set (SS), comprising all randomly assigned participants who received ≥1 dose of study treatment, according to the study treatment received. The level of statistical significance was set at 0.05 (2-sided). No adjustment for multiple testing was performed.

Efficacy analysis

Statistical analyses were performed blinded to treatment allocations according to a prespecified statistical analysis plan (SAP) (Supplementary material Appendix A) using Stata/SE, version 17.0 (StataCorp). For continuous variables, means and SDs are presented, as well as frequencies and percentages for categorical variables. Both unadjusted and covariate adjusted analyses were performed for each outcome, with conclusions drawn from the adjusted analyses. Covariates were considered based on their potential to influence outcome variables. The covariates adjusted for included stratification variables [study site, age in years (continues variable)], sexual activity 28 d prebaseline, UTI history (number of UTIs reported in last 12 mo), and, where applicable, the baseline measure for a continuous outcome in the same scale as the outcome. Adjustment for stratification variables is recommended by the Committee for Proprietary Medicinal Products [26] as stratification creates positive correlations between treatment groups, and failing to adjust for these variables results in upwardly biased SE estimates for the treatment effect. This, in turn, will lead to wide 95% CI, low type 1 error rates, and reduced statistical power [27]. Age (mostly young and postmenopausal women), sexual intercourse, and UTI history are factors shown to be associated with UTI incidence [[28], [29], [30]]. For continuously measured outcomes, residual distributions were examined, and if assumptions were found to be violated, the variables were log-transformed. For other regression models used, all assumptions were checked and appeared reasonable. Checks included inspection of Martingale residuals for Cox regression and visual inspection of the relationship between continuous covariates and the outcome on the log-scale. Secondary analyses of the primary and other UTI-related outcomes included age-squared as an additional adjustment factor to allow for a nonlinear effect of age on UTI risk.

The primary analysis of the primary outcome, incidence of culture-confirmed UTIs at any time during the 6-mo intervention period, was conducted using log Poisson regression with robust variance due to convergence issues with log binomial regression model that was prespecified in the SAP. Effect estimates are presented as RR and 95% CI for culture-confirmed UTI in the whole cranberry powder compared with the placebo group. Secondary outcomes involving incidence of UTI and/or UTI symptoms were analyzed using log binomial regression or log Poisson regression with robust variance in the case of nonconvergence. For outcomes with low event rates, Fisher’s exact test was conducted without covariate adjustment. Analysis of time from random assignment to first culture-confirmed UTI episode included only participants with ≥1 culture-confirmed UTIs during the 6-mo intervention and was assessed using Cox proportional hazards regression. Effect estimates are presented as hazard ratio (HR) and 95% CI for whole cranberry powder compared with placebo. Total count of culture-confirmed UTIs per participant over the 6-mo intervention period between groups was analyzed using negative binomial regression with effect estimates presented as incidence rate ratio (IRR) and 95% CI for whole cranberry powder compared with placebo. Deviations from the SAP and post hoc analyses are described in Supplementary material.

Safety analysis

For outcomes defined as presence or absence of a condition (e.g., AEs, SAEs, and physical examination), differences between treatment groups were analyzed using exact binomial tests or Fisher’s exact test. Effect estimates are expressed as risk difference calculated as risk in whole cranberry powder minus risk in placebo. Differences between groups in changes in parameters from baseline at 6 mo for continuously measured outcomes (vital signs, blood biochemistry, hematology, and some urinalysis variables) were analyzed using analysis of covariance, adjusted for covariates as above, and the corresponding baseline variable. The unadjusted model included the corresponding parameter at baseline only. Effect estimates are expressed as mean difference (95% CI). Differences between randomly divided groups for categorical variables (some urinalysis variables) were analyzed using ordinal regression model adjusted as detailed above with effect estimates expressed as odds ratio (OR) (95% CI) of higher compared with lower categories for whole cranberry powder compared with placebo at 6 mo.

Sensitivity analysis

In the primary analysis, participants who were lost to follow-up or discontinued the intervention and recorded 0 culture-confirmed UTIs were assumed to not have experienced a culture-confirmed UTI during the study period. To assess the sensitivity of results to this assumption, the analyses were repeated, first by including these participants as UTI-positive in the analysis and second by excluding these participants from the analyses.

Results

A total of 150 females were randomly assigned: n = 75 to the whole cranberry and n = 75 to the placebo groups (Figure 1). Of these, 14 discontinued the intervention or were lost to follow-up (9% dropout rate), 3 from the whole cranberry, and 11 from the placebo group. Of the 150 randomly assigned participants, 5 were excluded from the FAS because either no intervention product was taken or no postbaseline data were collected (2 in the whole cranberry and 3 in the placebo groups) as per the SAP. Therefore, the FAS comprised n = 145 participants (73 in the whole cranberry and 72 in the placebo groups). Additional information about the study population can be found in Supplementary materials (Supplementary Results). The treatment groups appeared balanced with respect to the majority of reported baseline characteristics (Table 1). Notably, the cranberry group included more postmenopausal females, and the first UTIs occurred at a later age compared with the placebo group. More than 95% and 88% of participants in the whole cranberry and placebo groups, respectively, consumed ≥80% of prescribed study capsules over the 6-mo study period, and >80% of participants were compliant overall (Supplemental Table 2).

FIGURE 1.

Flow diagram of participants through the study. FAS, full analysis set; PPS, per-protocol set; SS, safety set

TABLE 1.

Baseline characteristics of all randomly assigned participants.

	Placebo	Whole cranberry powder	Total
	n = 75	n = 75	n= 150
Trial site
Adelaide, South Australia	37	37	74
Sydney, New South Wales	30	30	60
Brisbane, Queensland	3	4	7
Gold Coast, Queensland	5	3	8
Melbourne, Victoria	0	1	1
Age (y)	34.2 (14.0)	35.1 (14.9)	34.6 (14.4)
Age category (n, %)
18–24 y	30 (40.0)	27 (36.0)	57 (38.0)
25–34 y	17 (22.7)	19 (25.3)	36 (24.0)
35–44 y	8 (10.7)	7 (9.3)	15 (10.0)
≥ 44 y	20 (26.7)	22 (29. 3)	42 (28.0)
Postmenopausal (n, %)	11 (14.7)	16 (21.3)	27 (18.0)
Height (cm)	166.1 (7.0)	166.6 (5.9)	166.3 (6.4)
Body weight (kg)	65.7 (11.0)	68.6 (13.9)	67.1 (12.6)
BMI (kg/m2)	23.8 (3.8)	24.7 (4.9)1	24.3 (4.4)
Systolic blood pressure (mmHg)	112.3 (12.7)	114.1 (12.1)	113.2 (12.4)
Diastolic blood pressure (mmHg)	72.1 (10.2)	73.5 (8.8)	72.8 (9.5)
Sexually active 28 d prebaseline (n, %)	61 (81.3)	59 (78.7)	120 (80.0)
Number of UTIs reported in previous 6 mo	2.2 (0.8)	2.1 (0.8)	2.2 (0.8)
Number of UTIs reported in previous 6 mo (n, %)
0	1 (1.3)	1 (1.3)	2 (1.3)
1	8 (10.7)	13 (17.3)	21 (14.0)
2	47 (62.7)	41 (54.7)	88 (58.7)
3	14 (18.7)	17 (22.7)	31 (20.7)
4	4 (5.3)	2 (2.7)	6 (4.0)
5	1 (1.3)	1 (1.3)	2 (1.3)
Number of reported UTIs in previous 12 mo	3.3 (1.0)	3.3 (1.1)	3.3 (1.0)
Number of reported UTIs in previous 12 mo (n, %)
0	0 (0.0)	0 (0.0)	0 (0.0)
1	0 (0.0)	0 (0.0)	0 (0.0)
2	14 (18.7)	15 (20.0)	29 (19.3)
3	36 (48.0)	39 (52.0)	75 (50.0)
4	17 (22.7)	11 (14.7)	28 (18.7)
5	6 (8.0)	6 (8.0)	12 (8.0)
6	2 (2.7)	2 (2.7)	4 (2.7)
7	0 (0.0)	2 (2.7)	2 (1.3)
Risk factors for UTI development
Age at first UTI (y)	18.9 (9.5)	19.6 (6.7)	19.3 (8.2)
Age category at first UTI (n, %)
<18 y	21 (44.7)	14 (31.1)	35 (38.0)
18–24 y	20 (42.6)	26 (57.8)	46 (50.0)
25–34 y	2 (4.3)	3 (6.7)	5 (5.4)
35–44 y	2 (4.3)	1 (2.2)	3 (3.3)
≥ 44 y	2 (4.3)	1 (2.2)	3 (3.3)
Maternal history of UTIs (n, %)
Yes	24 (32.0)	30 (40.0)	54 (36.0)
No	36 (48.0)	29 (38.7)	65 (43.3)
Unknown	15 (20.0)	16 (21.3)	31 (20.7)
New sexual partner in the last year (n, %)
Yes	26 (34.7)	18 (24.0)	44 (29.3)
No	46 (61.3)	56 (74.7)	102 (68.0)
Declined	3 (4.0)	1 (1.3)	4 (2.7)

Abbreviation: UTI, urinary tract infection.

All values are mean (SD) or frequency and percentages.

¹One participant in the whole cranberry powder group had a BMI exceeding the inclusion criteria of >35 kg/m². As this participant was suitably enrolled before the introduction of the BMI inclusion criteria, the participant remained in the trial and was included in the full analysis.

UTI results are summarized in Table 2. Adjusted risk for UTI incidence was 52% lower in the whole cranberry powder group than the placebo group (adjusted RR: 0.48; 95% CI: 0.26, 0.87; P = 0.01, adjustment variables included age at baseline [continuous variable], sexual activity, UTI history at baseline [number of UTIs reported in last 12 mo] and site). Escherichia coli was the most frequent uropathogenic species detected for UTIs, and its incidence at any time during the 6-mo period was lower in the whole cranberry powder group than the placebo group by 51% (adjusted RR: 0.49; 95% CI: 0.24, 1.01; P = 0.05). Incidence of UTIs with urinary frequency or urgency symptomatology (adjusted RR: 0.29; 95% CI: 0.13, 0.63; P < 0.01) and any one or more culture-confirmed symptom associated with UTI (adjusted RR: 0.42; 95% CI: 0.23, 0.78; P = 0.01) were also reduced in the whole cranberry powder group compared with the placebo group. Significant differences between groups in incidence of symptomatic suspected UTI (adjusted RR: 0.66; 95% CI: 0.41, 1.05; P = 0.08) and culture-confirmed dysuria (adjusted RR: 0.57; 95% CI: 0.29, 1.11; P = 0.10) were not observed. Event rates for other symptomatic outcomes were too low to be able to draw any conclusions. Time to first UTI incident was of longer duration for participants in the whole cranberry powder group than those in the placebo group (adjusted HR: 0.36; 95% CI: 0.18, 0.74; P = 0.01). In addition, the mean total number of UTIs per participant recorded over the 6-mo intervention period was lower in the whole cranberry powder group than the placebo group (adjusted IRR: 0.41; 95% CI: 0.21, 0.79; P = 0.01). Sensitivity analyses (Supplemental Tables 3–5) showed that the direction, magnitude, and significance of the estimated treatment effects on reducing the incidence of culture-confirmed UTIs remained consistent when various assumptions about the UTI incidence status of participants lost to follow-up were tested.

TABLE 2.

UTI-related outcomes over the 6-mo study period for the FAS.

	Placebo no./total no. (%)	Whole cranberry powder no./total no. (%)	Unadjusted RR (95% CI)1	P value	Adjusted RR (95% CI)1	P value
Incidence of culture-confirmed UTI (>108 cfu/L) at any time during the 6-mo intervention period [primary outcome]2	25/72 (34.7)	13/73 (17.8)	0.51 (0.29, 0.92)	0.03	0.48 (0.26, 0.87)	0.01
Age-squared term included in adjusted model2					0.46 (0.25, 0.82)	0.01
Incidence of symptomatic suspected3 UTI at any time during the 6-mo intervention period2	28/72 (38.9)	20/73 (27.4)	0.70 (0.44, 1.13)	0.15	0.66 (0.41, 1.05)	0.08
Incidence of culture-confirmed dysuria at any time during the 6-mo intervention period2	18/72 (25.0)	12/73 (16.4)	0.66 (0.34, 1.26)	0.21	0.57 (0.29, 1.11)	0.10
Incidence of culture-confirmed urinary frequency or urgency at any time during the 6-mo intervention period	23/72 (31.9)	7/73 (9.6)	0.30 (0.14, 0.66)	<.01	0.29 (0.13, 0.63)	<.01
Incidence of culture-confirmed fever at any time during the 6-mo intervention period4	1/72 (1.4)	3/73 (4.1)	-	0.62	-	-
Incidence of culture-confirmed suprapubic pain at any time during the 6-mo intervention period4	8/72 (11.1)	6/73 (8.2)	-	0.59	-	-
Incidence of culture-confirmed macroscopic hematuria at any time during the 6-mo intervention period4	2/72 (2.8)	3/73 (4.1)	-	1.00	-	-
Incidence of any 1 or more culture-confirmed symptoms associated with UTI at any time during the 6-mo intervention period2	25/72 (34.7)	12/73 (16.4)	0.47 (0.26, 0.87)	0.02	0.42 (0.23, 0.78)	0.01
Presence of Escherichia coli on culture at any time of a culture-confirmed UTI during the 6-mo intervention period2,6	20/72 (27.8)9	11/73 (15.1)	0.54 (0.28, 1.05)	0.07	0.49 (0.24, 1.01)	0.05
Presence of Enterococci on culture at any time of a culture-confirmed UTI during the 6-mo intervention period4	1/72 (1.4)	0/73 (0.0)	-	0.50	-	-
Presence of Enterobacter spp. on culture at any time of a culture-positive UTI during the 6-mo intervention period4	1/72 (1.4)10	0/73 (0.0)	-	0.50	-	-
Presence of Staphylococcus saprophyticus on culture at any time of a culture-confirmed UTI during the 6-mo intervention period4	2/72 (2.8%)9	0/73 (0.0)	-	0.24	-	-
Presence of other uropathogen5 on culture at any time of a culture-positive UTI during the 6-mo intervention period4	3/72 (4.2%)10	2/73 (2.7)	-	0.68	-	-
	Mean (SD)	Mean (SD)	Unadjusted HR (95% CI)7	P value	Adjusted HR (95% CI)7	P value
Time (d) to first culture-confirmed UTI episode over the 6-mo intervention period	69.7 (52.3)	91.9 (39.1)	0.44 (0.23, 0.87)	0.02	0.36 (0.18, 0.74)	0.01
	Mean (SD)	Mean (SD)	Unadjusted IRR (95% CI)8	P value	Adjusted IRR (95% CI)8	P value
Total number of culture-confirmed UTIs per participant recorded over the 6-mo intervention period9	0.5 (0.8)	0.2 (0.6)	0.48 (0.25, 0.91)	0.02	0.41 (0.21, 0.79)	0.01
Number of UTIs reported over the 6-mo intervention period	no./total no. (%)	no./total no. (%)
0	47/72 (65.3)	60/73 (82.2)	-	-	-	-
1	18/72 (25.0)	8/73 (11.0)	-	-	-	-
2	5/72 (6.9)	5/73 (6.8)	-	-	-	-
3	2/72 (2.8)	0/73 (0.0)	-	-	-	-

Abbreviations: CI, confidence interval; HR, Hazard ratio; IRR, Incidence rate ratio; RR, relative risk; UTI, urinary tract infection.

¹RR from log binomial regression model. Adjusted RR model included age at baseline (continues variable), sexual activity, UTI history at baseline (number of UTIs reported in last 12 mo) and site. Secondary analysis included addition of age-squared as covariate to allow for a potential nonlinear effect of age on UTI risk.

²Due to convergence issues, adjusted treatment effect estimate derived from log Poisson regression model with robust variance.

³Symptomatic suspected UTI was defined as a UTI-symptom-associated unscheduled visit.

⁴Fisher's Exact test conducted due to low event rate. Adjusted analysis not conducted.

⁵Other microorganisms may include Klebsiella pneumoniae, Coagulase negative staphylococci, Candida Species, Enterococcus faecalis, Klebsiella oxytoca.

⁶Presence of Escherichia coli on culture at baseline was also included as an adjustment factor in adjusted analysis.

⁷HR from Cox proportional hazard regression. Adjusted HR model included age at baseline, age-squared, sexual activity, UTI history at baseline (number of UTIs reported in last 12 mo), and site. The reported mean time to first UTI is based on observed data and includes only those participants with incidence of ≥1 culture-confirmed UTI during the 6-mo intervention period (placebo group n = 25; whole cranberry powder group n = 13).

⁸IRR from negative binomial regression. Adjusted IRR model included age at baseline, age-squared, sexual activity, UTI history at baseline (number of UTIs reported in last 12 mo), and site. Placebo n = 72; whole cranberry powder, n = 73.

⁹One participant in the placebo group presented with Escherichia coli and Staphylococcus saprophyticus on 2 different occasions.

¹⁰One participant in the placebo group presented with Enterobacter spp. and Klebsiella oxytoca on 2 different occasions.

Results for the per-protocol UTI outcomes are summarized in Supplemental Table 6. The adjusted RR reduction in the whole cranberry powder compared with the placebo group was 40% (adjusted RR: 0.60; 95% CI: 0.32, 1.14; P = 0.12), and when age-squared was included in the model 46% (adjusted RR: 0.54; 95% CI: 0.28, 1.06; P = 0.08), but neither of the models was statistically significant. Incidence of culture-confirmed urinary frequency or urgency was reduced in the whole cranberry powder group compared with the placebo group by 65% (adjusted RR: 0.35; 95% CI: 0.14, 0.86; P = 0.02). The incidence of any one or more culture-confirmed symptoms associated with UTI was reduced by 50% (adjusted RR: 0.50; 95% CI: 0.25, 1.00; P = 0.05). Other UTI outcomes were not statistically significantly different between groups, including time to first culture-confirmed UTI episode.

Other UTI-related results are summarized in Supplemental Table 7–10. Sexual health outcomes are reported in Supplemental Tables 11 and 12. In brief, bothersome scores for female sexual matters associated with UTI symptoms did not differ between treatment groups (see Supplemental Results for further description of sexual health outcomes).

There were no differences between the whole cranberry powder group compared with the placebo group for most safety outcomes, including vital signs (Supplemental Tables 13 and 14), incidence of unsatisfactory physical examinations (Supplemental Table 15), blood hematology and serum biochemistry outcomes (Supplemental Tables 16 and 17) and urinalysis outcomes (Supplemental Tables 18 and 19). The only differences that occurred were significant increases in serum globulin and serum alkaline phosphatase concentrations in the whole cranberry powder group compared with the placebo group, but these differences were small and not of clinical relevance (Supplemental Table 17). The incidence of most SAEs (n = 1) and AEs was approximately equal across treatment groups (Supplemental Tables 20 and 21, respectively). As expected, urinary related AEs (dysuria, pollakiuria, micturition urgency, and suprapubic pain) were detected more frequently in the placebo than the whole cranberry powder group (Supplemental Table 21). Incidence of treatment-related AEs was generally low (Supplemental Table 22). The risk of SAEs during the study period was generally low and did not differ between groups (Supplemental Table 23). No severe treatment-related AEs were reported. See Supplementary Results for further description of safety outcomes.

Discussion

This double-blind, randomized, placebo-controlled 6-mo clinical intervention trial in females with rUTI demonstrated that daily consumption of whole cranberry fruit powder reduced the risk of culture-confirmed UTI compared with placebo. Consistent findings were also observed for several UTI-related secondary outcomes, and the supplement was reported to be safe for consumption.

The results from the current study are similar to previous randomized controlled trials and meta-analyses of these trials, which investigated the effects of cranberry in various forms compared with placebo/control on UTI incidence in females with rUTI. Although the effects from these studies differed in magnitude and level of statistical significance, the direction of effects was generally toward protection against rUTI [18,21,23,[30], [31], [32], [33], [34], [35]], with the exception of Barbosa-Cesnik, which showed higher rUTI rates in the cranberry than the placebo group [36]. This latter study differed from other studies in that they enrolled females presenting at clinic with acute UTI rather than a history of rUTI, used a lower bacteria threshold to define UTI (10³ cfu/mL), and additionally used juice as the placebo, which although it did not contain cranberry, may have inadvertently contained similar active ingredients to those found in cranberry juice such as ascorbic acid [36]. Studies that enrolled mixed sex or pooled participants with complicated and uncomplicated UTIs generally reported no effect of cranberry products compared with placebo, likely due to complexities associated with UTI in males and complicated UTIs requiring different treatments [37,38].

The criteria for defining rUTI differed across studies. Although the current study defined rUTI as ≥2 UTIs in the past 6 mo or ≥3 UTIs in the past year as recommended [[39], [40], [41]], others used lower rates of ≥1 in the past 6 mo or ≥2 in the past year [21,32,34]. Despite differences in inclusion criteria all studies reported reduced rates of UTI with cranberry products compared with placebo [21,32,34]. However, results may be sensitive to the exclusion of participants with a history of higher number of UTIs (>5 UTIs in the past 6 or 12 mo). Babar et al. [23] used similar inclusion criteria as the current study (≥2 UTIs in the past 6 mo or ≥3 UTIs in the past year) and did not show an overall significant reduction in UTI incidence. However, they did detect a significant reduction in a subgroup of females with a history of <5 UTIs in the preceding year [23]. Thus, excluding participants with >5 UTIs in the past 6 mo, as in the current study, may be an important criterion.

Although the current study reported reduced incidence in culture-confirmed UTI with whole cranberry powder compared with placebo (RR: 0.48; 95% CI: 0.26, 0.87; P = 0.01), a reduction in incidence of symptomatic suspected UTI was not detected (RR: 0.66; 95% CI: 0.41, 1.05; P = 0.08). In contrast, Maki et al. [32], the largest study to date (n = 373), showed a reduced incidence of investigator-diagnosed symptomatic UTIs (IR: 0.61: 95% CI: 0.41, 0.91; P = 0.02, adjusted for antibiotic use) and symptomatic UTIs by clinical diagnosis with pyuria (IRR: 0.63; 95% CI: 0.40, 0.97; P = 0.04, adjusted for antibiotic use) with cranberry juice compared with placebo, but no difference between treatment groups for culture-confirmed UTIs (HR: 0.97; 95% CI: 0.56, 1.67; P = 0.91). The reason for the discrepancy in culture-confirmed incidence is not clear, but differences in symptomatic UTIs may be explained by the method in which the diagnoses were made. In Maki et al. [32], UTIs were clinically diagnosed by investigators, whereas in the current study, symptomatic UTI was based on participant-reported UTI symptoms and not on a clinical diagnosis. It is, therefore, possible that in the current study, the symptomatic suspected UTI outcome included false-positive cases for symptomatic suspected UTIs and, in combination with the small sample size, resulted in a nonsignificant effect (type II error).

There were no safety concerns related to the cranberry treatment reported in the current study, and the dropout rate was low at 9%. Jepson et al. [19] reported in their meta-analysis on cranberry products and UTI that many studies reported low compliance and high withdrawal/dropout problems attributable to palatability/acceptability of the products, primarily cranberry juice. Therefore, it can be suggested that a juice or beverage product may be less acceptable for use as a long-term prophylactic treatment, and a supplement form, such as the supplement used in the current trial, may be more suitable.

Strengths of the current study included targeting females at elevated risk for UTI episodes, who were appropriately clinically diagnosed with a history of rUTI, but otherwise healthy and most likely to benefit from cranberry supplementation. Females experiencing higher numbers of UTIs (>5 UTIs in the past 6 mo) were excluded as it was believed that they may not respond to supplement treatment. The fact that incidence rate of UTI during the study in the placebo group (34%) was as predicted in sample size estimates (25%–35%) supports that participants were appropriately diagnosed with rUTI and that the methods for diagnosing UTIs during the study were suitable and effective. Finally, sensitivity analyses were conducted and showed that the estimated treatment effects were unchanged when testing various assumptions about the UTI incidence status of participants lost to follow-up.

The limitations of the current study included, despite using an objective and standard measure to define UTIs, namely microbial growth on urine culture of >10⁸ cfu/L, it is likely that we missed some positive incidences of UTI over the intervention period. It has been reported that most females who present with typical UTI complaints but have a negative culture still do have an undetectable Escherichia coli infection [42]. Future studies may wish to consider clinical assessments that include a combination of outputs, such as urinary culture results, urinalysis (pyuria), and other common symptoms, to provide a more complete diagnosis of UTI. Although the initial target sample size of n = 300 was not achieved (increasing the type II but not affecting the type I error risk), the absolute reduction in culture-confirmed incidence observed between groups (17%) was greater than initially predicted (15%), which may explain significant results despite a reduced sample size of n = 150. Future studies should record and include as a confounder in the analyses the time since the last UTI before participant enrollment. Although microbial growth on urine culture (≥10⁷ cfu/L [10⁴ cfu/mL]) 7 d prebaseline was an exclusion criterion, Maki et al. [30] showed that recent UTI, <90 d prior to screening, was an important variable associated with the odds of having ≥1 UTI. The prespecified primary analysis was conducted on the FAS (a sample of 145 out of 150 participants), which is as complete and as close as possible to the intention-to-treat ideal of including all randomized participants [24]. However, as this is a modification from the intention-to-treat principle, potential bias in the estimation of the treatment effect needs to be considered. Jakobsen et al. [25], however, noted that when the fraction of individuals excluded is small, here 5 of 150 (3.3%), and those missing are unlikely to be individuals likely to be at greatest or lowest risk of outcome, then the risk of bias is negligible. Furthermore, sensitivity analyses testing several plausible assumptions about the outcome for participants excluded from the FAS resulted in the same substantive conclusions as the primary analysis of the primary outcome (see Supplemental materials, Tables 3–5). The per-protocol analysis, which included only n = 117 (out of 150), is at greater risk of bias and resulted in slightly smaller treatment estimates (See Supplemental Discussion for further information). Additionally, this study reports on a substantial number of secondary outcomes related to UTI incidence. These were not prespecified as confirmatory secondary analyses and were not adjusted for multiple testing, and as such, care must be taken when interpreting the reported P values. More dropouts were reported in the placebo group (n = 11) compared with the cranberry (n = 3) group. Reasons for withdrawal did not suggest a relationship to treatment allocation. Finally, it is important to note that the results are generalizable only to females with uncomplicated rUTI, excluding severe cases, over a 6-mo period.

In conclusion, the present study provides robust evidence that among females with recurrent UTI, consumption of whole cranberry fruit powder (500 mg/d Pacran) capsules, compared with placebo, does not impact safety markers and reduces the incidence of culture-confirmed UTI supported by improvements in several other UTI-related outcomes.

Author contributions

The authors’ responsibilities were as follows – WS, BBE: designed research in collaboration with the sponsor and the raw material supplier; WS, BBE: responsible/overseeing the execution of the intervention; JB, AV: developed the statistical analysis plan and performed statistical analysis; WS: drafted the first version of the manuscript and had primary responsibility for final content; BBE, JB, AV: contributed to writing of the manuscript and all authors: read and approved the final manuscript.

Funding

Swisse Wellness Pty Ltd was the trial sponsor and Givaudan Flavors Corp was the raw material supplier.

Data availability

Data described in the manuscript, code book, and analytic code will be made available upon request pending application and approval. Applications can be made to the corresponding author.

The full trial protocol can be accessed by emailing the corresponding author.

Disclaimers

The funding source and the raw material supplier, in collaboration with the research scientists, designed the trial and monitored its implementation, but had no influence over the analyses, reporting, interpretation of the data and preparation of the manuscript. The manuscript was reviewed by the sponsor and the raw material supplier before the submission to the journal, but they had no influence over the manuscript content.

The authors have prepared the manuscript according to the CONSORT 2010 statement.

Conflict of interest

Financial sponsorship for the study was provided by Swisse Wellness Pty Ltd to the Commonwealth Scientific Industrial Research Organisation. All authors report no conflicts of interest.

Appendix A. Supplementary data

The following are the Supplementary data to this article: