Daily Cranberry Juice for the Prevention of Asymptomatic Bacteriuria in Pregnancy: A Randomized, Controlled Pilot Study

Materials and Methods 

Results 

From July 2005 through July 2007 a total of 188 women were enrolled in this pilot investigation (see figure). There were no differences in demographic characteristics (table 1). There were 27 UTIs in this cohort. There was 1 case of Enterobacter faecalis cystitis in a woman in group B and 3 cases of pyelonephritis due to Escherichia coli (2 in group A, 1 in group B), with the remainder of cases attributed to ASB (table 2). Five women had more than 1 UTI. There was a trend toward fewer UTIs, asymptomatic and symptomatic, in those women who received multiple daily doses of CJC compared to those who received placebo. This trend persisted with single daily dosing of CJC although the magnitude of the difference was less (table 3).

View Large Image Download to PowerPoint

Participant flow chart for cranberry for prevention of asymptomatic bacteriuria in pregnancy

Table 1. Demographics by group

		Group A	Group B	Group C	p Value⁎
	No. women	58	67	63
	Mean maternal age ± SD	25.8±5.6	27.7±5.4	25.6±5.0	0.058
	No. gravida (%):
	1	16(27.6)	12(17.9)	27(42.9)
	2	14(24.1)	21(31.3)	15(23.8)
	3 or More	28(48.3)	34(50.8)	21(33.3)	0.034
	No. para (%):
	0	22(37.9)	18(26.9)	30(47.6)
	1	17(29.3)	23(34.3)	22(34.9)
	2 or More	19(32.8)	26(38.8)	11(17.5)	0.053
	No. ethnicity (%):†
	NonHispanic white	11(18.9)	11(16.4)	10(15.9)
	Hispanic white	40(68.9)	45(67.2)	47(74.6)
	NonHispanic black	3(5.2)	6(8.9)	4(6.4)
	Asian/other	4(6.9)	5(7.5)	2(3.2)	0.89
	No. enrollment site (%):
	LBMMC	23(39.7)	28(41.8)	26(41.3)
	Santa Ana (UCI)	25(43.1)	24(35.8)	27(42.9)
	Manchester (UCI)	10(17.2)	15(22.4)	10(15.9)	0.84
	No. insurance status (%):
	Government	44(75.9)	53(79.1)	50(79.4)
	Private/other	14(24.1)	14(20.9)	13(20.6)	0.88
	No. yrs school (%):
	0–8	8(13.8)	8(11.9)	4(6.3)
	9–12	33(56.9)	41(61.2)	40(63.5)
	13 or More	17(23.3)	18(26.9)	19(13.2)	0.72
	No. employment status (%):
	Employed	25(41.4)	27(40.3)	26(41.3)
	Not employed	34(58.6)	40(59.7)	37(58.7)	0.99
	No. history of prior UTI (%):
	Yes	17(29.3)	18(26.9)	20(31.8)
	No	41(70.7)	49(73.1)	43(68.2)	0.83

⁎ ANOVA for continuous variables, Fisher's exact or chi-square test for categorical variables. † There were no Hispanic black subjects in the study population.

Table 2. Types of urinary tract infections with uropathogens

		Total UTIs	No. Group A	No. Group B	No. Group C
	Asymptomatic bacteriuria:
	E. coli	9	1⁎	3†‡	3
	Citrobacter freundii	5			1§
	Group B streptococcus	3	1	2
	Proteus mirabilis	3	1∥		1
	Other	3		1	2
	Total	23
	Symptomatic bacteriuria (cystitis):
	Enterobacter cloacae	1		1
	Total	1
	Pyelonephritis:
	E. coli	3	2⁎	1‡
	Total	3
	Total subjects with UTIs (%)		4 (6.9)	7 (10.4)	7 (11.1)

⁎ E. coli pyelonephritis also developed in 1 subject with E. coli ASB. † One subject with 2 bouts of E. coli ASB. ‡ E. coli pyelonephritis also developed in 1 subject with 2 bouts of E. coli ASB. § One subject with 5 bouts of C. freundii ASB. ∥ One subject with 2 bouts of P. mirabilis ASB.

Table 3. Incidence rate ratios for cases of asymptomatic bacteriuria and UTIs (intent to treat)

		IRR (95% CI)⁎
		Asymptomatic Bacteriuria	All UTIs
	Group A	0.43(0.14–1.39)	0.59(0.22–1.60)
	Group B	0.85(0.34–2.08)	0.85(0.36–2.01)
	Group C	1.0	1.0

⁎ Poisson.

More women in the once daily dosing group and the placebo group were likely to have at least 1 UTI during the study compared to the multiple daily dosing group (7 of 67 [10.4%] and 7 of 63 [11.1%] vs 4 of 58 [6.9%], respectively, Fisher's exact test p = 0.71). Similar results were seen in evaluating only those UTIs due to enteric bacteria (5 of 67 [7.5%], 5 of 63 [7.9%] and 3 of 58 [5.2%], respectively, p = 0.83), and the trend toward reduction in ASB alone persisted for multiple daily cranberry juice cocktail dosing (IRR 0.43, 95% CI 0.14–1.39) as well as for single daily cranberry juice cocktail dosing (IRR 0.85, 95% CI 0.34–2.08).

Compliance and tolerability were considerable obstacles during this investigation. Despite the change from 3 times daily to twice daily dosing actual dosing regimens did not differ among groups with 50.7% (34 of 67) of group A, 39.7% (23 of 58) of group B and 55.5% (35 of 63) of group C consuming placebo juice once or twice daily, p = 0.45. Compliance rates differed among groups with total doses prescribed for the duration of participation consumed at 65.7% ± 30.9% in group A, 78.7% ± 29.2% in group B and 76.9% ± 24.9% in group C, p = 0.03. Of 188 subjects 73 (38.8%) could not complete the study and withdrew, most for gastrointestinal upset including nausea, vomiting, diarrhea and dislike of taste (44 of 73). There were fewer withdrawals after the dose change was made (50 of 136, 36.8%) vs before (23 of 52, 44.2%, p = 0.35).

Evaluating the cohort on an intent to treat basis the median number of days in study was 152.5 (IQR 56 to 183) for group A, 158 (IQR 61 to 181) for group B and 171 (IQR 76 to 185) for group C, p = 0.26. For those who completed the study protocol the median number of days in study was 183 (IQR 161 to 195) for group A (41), 177 (IQR 165 to 185) for group B (31) and 182 (IQR 169 to 192) for group C (43), p = 0.29. For those who withdrew from study the median number of days in study was 56 (IQR 21 to 77) for group A (27), 56 (IQR 30 to 90) for group B (26) and 55.5 (IQR 28.5 to 80) for group C (20), p = 0.85. There were no differences between the groups with regard to obstetric or neonatal outcomes (table 4). No preterm deliveries at less than 34 weeks occurred in women with UTIs during this investigation.

Table 4. Obstetric and neonatal outcomes

		Group A	Group B	Group C	p Value⁎
	Mean gestational age (wks) at delivery ± SD (No.)	38.7±3.0 (55)	38.2±3.6 (58)	38.8±2.5 (57)	0.027
	No. preterm delivery less than 37 wks (%):
	Yes	6 (10.9)	11 (19.0)	4 (7.0)
	No	49 (89.1)	47 (81.0)	53 (93.0)	0.15
	No. preterm delivery less than 34 wks (%):
	Yes	2 (3.6)	4 (6.9)	2 (3.5)
	No	53 (96.4)	54 (93.1)	55 (96.5)	0.73
	No. route of delivery (%):
	Spontaneous vaginal delivery	36 (66.7)	41 (70.7)	40 (70.2)
	Instrumented vaginal delivery	5 (9.3)	4 (6.9)	2 (3.5)
	Cesarean/cesarean hysterectomy	13 (24.0)	13 (22.4)	15 (26.3)	0.80
	Mean birth wt ± SD (gm)	3,270±522	3,296±591	3,423±644	0.31
	No. low birth wt (%):
	Yes	4 (6.9)	4 (6.0)	2 (3.2)
	No	54 (93.1)	63 (94.0)	61 (96.3)	0.72
	No. 1-min Apgar less than 7 (%):
	Yes	3 (5.5)	5 (8.8)	2 (3.5)
	No	50 (94.5)	52 (91.2)	55 (96.5)	0.52
	No. 5-min Apgar less than 9 (%):
	Yes	4 (7.5)	5 (8.8)	5 (8.8)
	No	49 (92.5)	52 (91.2)	52 (91.2)	1.0
	No. admission to neonatal intensive care unit (%):
	Yes	3 (5.7)	7 (11.9)	6 (10.5)
	No	50 (94.3)	52 (88.1)	48 (89.5)	0.51

⁎ ANOVA for continuous variables, Fisher's exact or chi-square test for categorical variables.

Discussion 

Our investigation provides support for a unique approach to the reduction of asymptomatic bacteriuria in pregnancy and its associated adverse perinatal outcomes. The standard of obstetric care remains screening for asymptomatic bacteriuria and treatment if the diagnosis is made.11 However, the current recommendations for screening may not apply to those women with poor compliance or late entry to prenatal care and do not address issues related to provider error or optimal posttreatment surveillance. The option to combine routine urinary screening with a nonharmful foodstuff, cranberry, to reduce the risk of gestational bacteriuria and its attendant potential complications is attractive for public health and cost considerations, especially when factoring in the expense of caring for a premature newborn and the subsequent costs related to the lifelong disability that often affects survivors of premature birth.

The mechanism by which cranberry may prevent urinary tract infection is unknown although there is a growing body of evidence that proanthocyanidins or condensed tannins, components in many berry products, inhibit the adhesion of piliated enteric bacteria such as E. coli to the uroepithelium.12, 13

In the 2001 Cochrane Library review of cranberry for the prevention and treatment of UTI the authors believed that there was preliminary evidence supporting its efficacy but that the published trials had major limitations including lack of control groups, small sample sizes, lack of controlled diets or dietary assessment, inappropriate analysis of data for dropouts or withdrawals, and lack of blinding.9 Importantly this Cochrane review noted that there were no investigations of the role of cranberry in preventing UTI in young patients or in pregnant women. This review included only 5 trials, all of which lacked a description of the product. Outcome measures were also varied with some researchers focusing on bacteriuria and pyuria, and others on symptomatic UTIs. The appropriate product, dose, duration of intervention and mechanism(s) of action were largely unknown or were not clearly elucidated. The same authors of the Cochrane review published updates in 200710, 14 and concluded that on meta-analysis of 4 high quality randomized controlled trials15, 16, 17, 18 cranberry products significantly reduced the incidence of symptomatic UTIs in 12 months (RR 0.66, 95% CI 0.47–0.92) compared with placebo or control, particularly in women with recurrent UTIs. These studies involve primarily women,15, 16 subjects with spinal cord injury17 or the elderly.18 To date to our knowledge no investigations on the effect of cranberry on urinary tract infection in pregnant women have been published.

Similar to our clinical trial withdrawals or losses to followup are significant in other published studies, as high as 47% or more.10, 18, 19 This has led some to suggest that drinking considerable amounts of cranberry juice during a long period such as the duration of pregnancy may not be acceptable.10 However, the possibility exists that different cranberry formulations such as capsules or tablets will have better efficacy and improved compliance. Stothers reported yearlong compliance rates of 70% to 100% with cranberry capsules, and lesser rates with cranberry juice in a trial evaluating the clinical and cost-effectiveness of cranberry for uroprotection.15 The nausea and vomiting during pregnancy certainly were poor prognosticators for compliance during this investigation, and any gastrointestinal symptoms related to intolerability of the juice or placebo may have exacerbated or been exacerbated by physiological changes in pregnancy. We altered the treatment regimens after approximately a third of the subjects were randomized due to concerns about compliance. There was a mild improvement in retention of compliant subjects in the investigation after this alteration was made. As there are in vitro data to suggest that the anti-adhesion activity of cranberry juice on fimbriated E. coli persists for 10 hours after ingestion, twice daily dosing of cranberry juice may be adequate for the prevention of UTIs in pregnancy.20

We acknowledge the limitations of this investigation including the small sample size and the lack of bioassay for compliance. As expected compliance and tolerability to the cranberry juice product were limitations, and resulted in a more than 30% dropout rate. As a result of these difficulties a change in the dosing regimens was required during the study. A minor weakness is that the provision of additional followup for those subjects with poor tolerance or compliance to the juice could have introduced bias in the ultimate clinical outcomes. Conclusive information about the benefits of daily cranberry juice ingestion can only be gleaned from larger clinical trials in which consistent treatments are applied. Additional evidence will soon be available from other NCCAM supported clinical trials in which the same cranberry juice and placebo products were used.