By means of our systematic search, we identified six trials[
12-17]. One was a protocol of an ongoing study[
15], so it was not included. Thus, eventually five trials, including a total of 377 subjects (194 in the experimental group and 183 in the control group), met our predefined inclusion criteria. The characteristics of the included studies are presented in Table . The list of excluded trials (
n = 22) is available upon request. The most usual reason for exclusion of a study was that the study was not randomised, the study was carried out in healthy volunteers, or the intervention was treatment with a symbiotic, not a probiotic alone. In addition, some studies were published in Japanese with no English abstract, and thus, were not accessible to the reviewers, even for initial screening.
All of the trials were full peer-reviewed publications. Four of the included studies were RCTs with a parallel design, and the remaining included RCT had a crossover design. All were placebo-controlled trials. The participants were adults (three RCTs,
n = 266) and children (two RCTs,
n = 111) with constipation defined as stated in Table . The following different probiotic strains were tested:
Bifidobacterium lactis DN-173 010,
E. coli Nissle 1917,
Lactobacillus casei rhamnosus Lcr35,
L. casei Shirota, and
L. rhamnosus GG. One RCT assessed the effectiveness of using
L. rhamnosus GG as an adjunct to lactulose therapy compared with treatment with lactulose alone[
17]. The durations of the interventions in the parallel-design studies were two weeks in one study, four weeks in two studies, and 12 wk in one study. The duration of the intervention in the crossover design study was eight weeks. The doses of the probiotic used ranged from 8 × 10
8 to 25 × 10
9 colony-forming units (CFU)/d.
Study description
RCTs in adults: The study by Mollenbrink and Bruckschen[
13] was a single-centre, randomised, double-blind, crossover trial that investigated the efficacy of treating 70 constipated patients with
E. coli Nissle 1917 or placebo. After four weeks of treatment, there was a significant difference in the mean number of stools per week in the
E. coli group compared with the placebo group (4.9 ± 1.5
vs 2.6 ± 1.0, respectively, MD 2.3 stools per week, 95% CI 1.7 to 2.9), which also remained significant at eight weeks (6 ± 1.3
vs 1.9 ± 1.5, respectively, MD 4.1, 95% CI 3.2 to 5). This study also revealed a significant difference between the probiotic and the control group in the incidence of hard stools (2/34
vs 16/30, respectively, RR 0.1, 95% CI 0.03 to 0.4). Both the effectiveness and tolerance of the treatment, as assessed both by a physician and the patients, were significantly better in those in the
E. coli group. The authors concluded that
E. coli Nissle 1917 is successful in the therapy of idiopathic chronic constipation.
The study by Koebnick et al[
12] was a single-centre, double-blind, placebo-controlled, randomised trial involving 70 patients with symptoms of chronic constipation. All of the patients received either a probiotic beverage containing
L. casei Shirota or placebo for four weeks. Patients completed a questionnaire related to their gastrointestinal symptoms, well-being, and stool habits, and underwent a medical examination weekly. The severity of constipation, flatulence, and bloating was divided into four categories (severe, moderately severe, mild, and no symptoms). Compared to the placebo group, those randomised to the
L. casei Shirota group experienced a significant improvement in the self-reported severity of constipation and stool consistency. That is, they experienced significant reductions in the occurrence of moderate and severe constipation (
P < 0.001), the degree of constipation (
P = 0.003), and the occurrence of hard stools (
P < 0.001), and increased their defecation frequency (
P = 0.004). However, the occurrence and degree of flatulence or bloating sensation did not significantly differ between the groups.
In the most recent study, Yang et al[
14] administered a fermented milk product containing
B. lactis DN-173 010 and some yoghurt strains (
S. thermophilus and
L. bulgaricus (1.2 × 10
9 CFU/pot 100 g) (experimental group) or an acidified milk containing non-living bacteria but no
B. lactis DN-173 010 or yoghurt strains (control group) for two weeks to constipated women. Comparison of the experimental group with the control group revealed a significantly higher stool frequency after one week of product administration (3.5 ± 1.5
vs 2.5 ± 0.9, respectively, MD 1.0 stool per week, 95% CI 0.6 to 1.4) and at two weeks (4.1 ± 1.7
vs 2.6 ± 1.0, respectively, MD 1.5 stool per week, 95% CI 0.7 to 1.6). The extent of defecation difficulty was assessed as 0-3 point defecation condition scores. In brief, 0 points indicates normal defecation, while 3 points indicates often abdominal pain or anal burning sensation to influence defecation. (Table for complete categorisation of defecation condition scores). Both at one and two weeks after product consumption, there was a significant improvement in the defecation condition scores in the experimental group compared with the control group: 1.1 ± 0.9
vs 1.6 ± 1.1, respectively (MD -0.5, 95% CI -0.85 to -0.18) at 1 wk and 0.8 ± 1.0
vs 1.6 ± 1.1, respectively (MD -0.79, 95% CI -1.14 to -0.44) at 2 wk. The stool consistency score was determined according to the Bristol Stool Scale. In brief, 0 points indicates stools like a sausage or a snake, smooth and soft, while 2 points indicates separating hard lumps, like fruit kernel (difficult discharge) (Table ). The stool consistency scores were significantly improved in the
B. lactis DN-173 010 group compared to the control group at 1 wk (1.0 ± 0.8
vs 1.4 ± 1.0, respectively, MD -0.4, 95% CI -0.73 to -0.12) and at 2 wk (0.6 ± 0.8
vs 1.3 ± 1.0, respectively, MD -0.7, 95% CI -1 to -0.4). There were no significant differences between groups in food intake and safety parameters. The researchers concluded that the administration of a fermented milk product containing
B. lactis DN-173 010 has a beneficial effect on stool frequency, defecation conditions, and stool consistency in adult women with constipation.
RCTs in children: Only two RCTs have addressed the use of probiotics in the treatment of constipation in children. In the study by
Banaszkiewicz and
Szajewska[
17], 84 children (aged: 2-16 years) with constipation (< 3 spontaneous bowel movements per week for at least 12 wk) were enrolled in a double-blind, randomised, placebo-controlled trial in which they received 1 mL/kg per day of 70% lactulose plus 10
9 CFU of
L. rhamnosus GG (experimental group,
n = 43) or a lactulose-containing placebo (control group,
n = 41) orally twice daily for 12 wk. The primary outcome measure was treatment success; all analyses were performed on an intention-to-treat basis. Treatment success was defined as ≥ 3 spontaneous bowel movements per week with no episodes of faecal soiling. Treatment success was similar in the control and experimental groups at 12 wk [28/41 (68%)
vs 31/43 (72%), respectively;
P = 0.7] and at 24 wk [27/41 (65%)
vs 27/43 (64%), respectively;
P = 1.0]. The groups also did not differ in their mean number of spontaneous bowel movements per week or episodes of faecal soiling per week at four, eight, and 12 wk. Adverse events and overall tolerance did not differ between groups. It was concluded that
L. rhamnosus GG, as dosed in this study, was not an effective adjunct to lactulose in treating constipation in children.
The study by Bu et al[
16] evaluated the effect of treating children with chronic constipation with
L. casei rhamnosus Lcr35 compared to magnesium oxide (MgO) or placebo; however, only the latter comparison is valid for this systematic review. For those treated with the probiotic (
n = 18) compared with placebo (
n = 9), the trial showed an increase in the treatment success defined as ≥ 3 spontaneous defecations per week with no episodes of faecal soiling (14/18
vs 1/9, respectively, RR 7, 95% CI 1.1 to 45), an increase in the defecation frequency (times/d) (0.57 ± 0.17
vs 0.37 ± 0.10, respectively, MD 0.2, 95% CI 0.1 to 0.3), a reduction in abdominal pain (frequency) (1.9 ± 1.6
vs 6.7 ± 3.3, respectively, MD -4.8, 95% CI -6.6 to -3), a reduction in the use of glycerin enemas during the four weeks of therapy (frequency) (1.6 ± 1.9
vs 4.0 ± 2.1, respectively, MD -2.4, 95% CI -4 to -0.8), and a decrease in the percentage of hard stools in the total number of defecations (22.4 ± 14.7
vs 75.5 ± 6.1, respectively, MD -53%, 95% CI -63 to -43). However, there was no difference between groups in the use of lactulose or the number of episodes of faecal soiling. No change in appetite was observed. However, the sample size was too small to draw any meaningful conclusion.
This article has been 