Table 2.
Study characteristics and data extraction
| Study | Mean age (SD) [range] unless otherwise stated | Participants | Setting | Study type | Intervention | Comparator | Duration of intervention | Outcome findings relevant to our systematic review (intervention vs. control) mean (SD) unless otherwise stated | GRADE certainty | |
|---|---|---|---|---|---|---|---|---|---|---|
| No | Characteristics | |||||||||
| Akbari et al. [24] |
Control: 82 (± 1.7) Intervention: 77.7 (± 2.6) |
60 | People with Alzheimer’s dementia | Welfare organisations, Iran | RCT |
PROBIOTIC (+ Milk) Lactobacillus acidophilus, Lactobacillus casei, Bifidobacterium bifidum, Lactobacillus fermentum (2 × 109 CFU/g of each) 200 ml/day |
Milk (200 ml/day) | 12 weeks |
Cognition: Improved MMSE with intervention + 27.90% (± 8.07) vs − 5.03% (± 3.00) (p < 0.001) |
Low |
| Hwang et al. [25] |
Control: 69.2 (± 7.0) Intervention: 68.0 (± 5.1) |
100 | Individuals with mild cognitive impairment | Community, South Korea | RCT |
PROBIOTIC Lactobacillus plantarum C29 (from kimchi) (≥ 1.25 × 1010 CFU/g) Fermented soybean (DW2009) 62.5% fermented soybean 800 mg/day |
Placebo capsules (cellulose) | 12 weeks |
Cognition: Improved combined cognitive function with intervention and specifically improved attention domain. No significant improvement in working or verbal memory domains Combine cognitive function: 0.95 to 0.79 vs. 0.68 to 0.65 (p for interaction = 0.02) Attention composite score: 1.44 to 1.05 vs. 0.82 to 0.70 (p for interaction = 0.02) Working memory composite score 1.12 to 1.08 vs. 0.98 to 1.06 (p for interaction = 0.63) Verbal memory composite score 0.99 to 0.98 vs. 0.87 to 0.93 (p for interaction = 0.21 |
Low |
| Kobayashi et al. [26] |
All: 82.5 (± 5.3) |
27 | Older persons attending monthly outpatient rehab with mild cognitive impairment | Community, Japan |
Pilot Single arm study |
PROBIOTIC Bifidobacterium breve A1 (~ 2 × 1010 CFU) Two capsules after meals |
Nil | 24 weeks |
Cognition: Improved cognition with intervention MMSE after 24 weeks + 1.7 (p < 0.01) Serial 7 s after 24 weeks + 1.1 (p < 0.01) ‘Resolution’ of MCI in 13/19 cases by 24 weeks DSST after 24 weeks + 0.74 (p = 3.2)—no significant change Mood: No change to mood No significant change to TMD scores and subscales |
Very low |
| Kotzampassi et al. [36] |
Control: 66.4 (± 11.9) Intervention: 65.9 (± 11.5) |
168 | All adult patients consecutively admitted for elective colonic resection with primary anastomosis | Surgical unit, Greece | RCT |
PROBIOTIC Lactobacillus acidophilus LA-5 (1.75 × 109 CFU), Lactobacillus plantarum (0.5 × 109 CFU), Bifidobacterium lactis BB-12 (1.75 × 109 CFU), Saccaromyces boulardii (1.5 × 109 CFU) Four capsules preoperatively then one capsule BD for next 14 days with 100 ml water |
Placebo capsules (powdered glucose polymer) | 14 days |
Length of stay: Reduced length of stay postoperatively with intervention Median time until discharge: 8 days vs. 10 days (log-rank: 20.30, p < 0.0001) |
Very low |
| Mañé et al. [20] |
Control: Median age: 69 [IQR: 66–82] Intervention 1: Median age: 70 [IQR: 67–83] Intervention 2: Median age: 71 [IQR: 65–84] |
50 | Older participants (> 65) institutionalised in one of two centres | Institutional geriatric centres, Spain | RCT |
PROBIOTIC + milk powder Lactobacillus plantarum CECT 7315 and CECT 7316 Low dose: 5 × 108 CFU/day in 20 g of powdered skimmed milk High dose: 5 × 109 CFU/day in 20 g of powdered skimmed milk |
Placebo (20 g of powdered skimmed milk) | 12 weeks |
Mortality: Trend towards positive effect on survival but underpowered study Mortality during treatment: low dose probiotic 0/13, High dose probiotic 0/19, Placebo 3/18 Physical function: No significant change in ADLs demonstrated No significant change in Barthel index (follow up data not shown) |
Low |
| McNaught et al. [32] |
Control: Median age: 71 [IQR: 65–77] [28–90] Intervention: Median age: 71 [IQR: 60–77] [28–87] |
103 | Adults admitted to ITU | Intensive Care Unit, UK | RCT |
PROBIOTIC + conventional therapy Proviva—oatmeal and fruit drink containing Lactobacillus plantarum 299v (5 × 107 CFU) 500 ml/day |
Conventional therapy alone (no placebo) | For duration of hospital admission or patient request to stop |
Mortality: No significant impact on mortality Mortality rate in both groups 35% |
Very low |
| Nomura et al. [35] |
Control: Median age: 69 [50–88] Intervention: Median age: 66 [30–83] |
64 | Patients scheduled for pancreaticoduodenectomy | Surgical centre, Japan | RCT |
PROBIOTIC BIO-THREE® Tablets Enterococcus faecalis T-110 2 mg, Clostridium butyricum TO-A 10 mg, Bacillus mesentericus TO-A 10 mg Six tablets/day |
Nil | Immediately after admission (3–15 days pre-operatively), restarted on second day post operatively and continued to discharge |
Mortality: Numbers too small to draw conclusions on mortality Only one death in study group (in control group due to abdominal bleeding) Length of stay: Reduced postoperative LOS in intervention group Median (range) days: 19 (11–40) vs. 24 (11–91) (p = 0.04) |
Very low |
| Rongrungruang et al. [33] |
Control: 68.9 (± 18.4) [20–97] Intervention: 73.1 (± 13.2) [30–94] |
150 | Ventilated patients | Intensive Care Unit, Thailand | RCT |
PROBIOTIC + standard care Lactobacillus casei—Yakult (8 × 109 CFU) 80 ml for oral care (after standard care) and additional 80 ml via enteral feeding OD |
Standard care (chlorhexidine 2% oral care QDS) | 28 days or until ETT removed |
Length of stay: No significant difference in LOS Median (range) days: 20 (2–106) vs. 19 (3–171) (p = 0.79) Mortality: No significant difference in mortality Day 28: 24% vs. 22.7% (p = 0.85) Day 90: 33.3% vs. 34.7% (p = 0.86) |
Very low |
| Shinkai et al. [16] |
Control: 70.9 (± 3.8) Intervention 1: (low dose) 71.0 (± 4.0) Intervention 2: (high dose) 70.8 (± 3.4) |
278 | Community dwelling older adults (> 65 years) | Community, Japan | RCT |
PROBIOTIC Heat killed Lactobacillus pentosus b240 Low dose: 2 × 109 cells High dose: 2 × 1010 cells One tablet/day at breakfast |
One tablet/day at breakfast | 20 weeks |
Mood: No improvement in mood with intervention Mental health, emotional role or social functioning subscales of SF-36 unchanged with intervention General health perception: 55.5(± 7.8) to 56.6 (± 8.2) vs. 54.9 (± 7.8) to 53.4 (± 8.6) Physical function: No improvement in physical function Self-reported physical functioning or physical role subscales of SF-36 not changed by intervention |
Moderate |
| Tamtaji et al. [28] |
Control: 78.8 (± 10.2) Intervention: 76.2 (± 8.1) |
79 | Older adults with a diagnosis of AD | Community, Iran | RCT |
PROBIOTIC + Selenium Lactobacillus acidophilus, Bifidobacterium bifidum, Bifidobacterium longum (2 × 109 CFU/day of each) + selenium 200 ng/day |
Selenium (exclude placebo from SR as not single intervention) | 12 weeks |
Cognition: Improved cognition with intervention MMSE: 1.5(± 1.3) vs 0.5(± 1.2) (p < 0.001) |
Moderate |
| Buigues et al. [17] |
Control: 73.4 (± 1.8) Intervention: 74.2 (± 1.6) |
50 | Mobile, non-demented nursing home residents | Nursing homes, Spain | RCT |
PREBIOTIC Darmocare Pre® (inulin min. 3375 mg, plus fructooligosaccharides min. 3488 mg per level measuring spoon of 7.5 g) After breakfast daily |
Placebo (maltodextrin) | 13 weeks |
Frailty: Overall no significant effect on frailty Improved exhaustion Mean frailty score (fried): Intervention: 2.8 (± 1.0) at baseline to 2.5 (± 0.8) after intervention Exhaustion (self-reported 0–3): 1.4 (± 1.7) to 0.8 (± 0.4) vs. 1.1 (± 1.7) to 1.7 (± 1.2) (p = 0.002) Cognition: No significant effect on MMSE MMSE: 26.5 (± 3.1) to 26.4 (± 2.2) vs. 26.1 (± 2.2) to 25.9 (± 2.1) (p = 0.85) Physical function: Improved hand grip strength No significant difference in ADLs (Barthel) or walking speed R hand grip strength: 10.6 (± 8.2) to 12.4 (± 3.2) vs. 11.5 (± 5.7) to 12.4 (± 3.2) (p < 0.05) Barthel index: 74.6 (± 17.7) to 77.1 (± 29.9) vs. 76.2 (± 13.0) to 78.3 (± 13.9) (p = 0.87) Walking speed (time needed to walk 4.6 m): 8.4 (± 6.0) to 7.9 (± 4.5) vs. 8.6 (± 9.0) to 8.7 (± 4.2) (p = 0.48) |
Low |
| Lewis et al. [31] |
Control: Median age: 74 [IQR: 63–80] Intervention: Median age: 76 [IQR: 65–84) |
142 | Consecutive inpatients with Clostridium difficile toxin assoc. diarrhoea | Hospital, UK | RCT |
PREBIOTIC + standard treatment Oligofructose 12 g/day |
Placebo (sucrose 12 g/day) + standard treatment | Started as soon as possible after dx and taken for 30 days after cessation of diarrhoea |
Mortality: No obvious significant difference in mortality demonstrated Died: 9 (13%) vs. 10 (14%) Length of stay: No obvious significant effect on LOS Median: 27 [IQR: 16–51] vs. Median: 29 [IQR: 14–58] |
Low |
| Theou et al. [22] |
Control: 75.9 (± 7.8) Intervention: 74.9 (± 6.9) |
50 | Mobile, non-demented nursing home residents | Nursing homes, Spain |
RCT 2ry analysis of [26] |
PREBIOTIC Darmocare Pre® (inulin min. 3375 mg, plus fructooligosaccharides min. 3488 mg per level measuring spoon of 7.5 g) 7.5 g after breakfast daily |
Placebo (maltodextrin) | 13 weeks |
Frailty: Intervention reduced frailty compared to placebo Significant interaction (p < 0.001) of time with the treatment group for FI 13 weeks follow up: Intervention group had lower FI levels compared to baseline (p < 0.001) vs. placebo had higher FI levels (p = 0.012) Average reduction of 0.02 (± 0.02) in intervention and increase of 0.01 (± 0.01) FI in placebo Moderately/severely frail group had more significant improvement |
Moderate |
| Jain et al. [30] |
Control: Median age: 73 [IQR: 65–80] Intervention: Median age: 72 [IQR: 62–77] |
90 |
ITU and HDU pts admitted for < 24 h No specific age range |
Intensive Care Unit, UK | RCT |
SYNBIOTIC Trevis™ capsules: Lactobacillus acidophilus La5, Bifidobacterium lactis Bb12, Streptococcus thermophilus and Lactobacillus bulgaricus (4 × 109 CFU of each) Raftilose™ powder, Orafti Active Food ingredients: Oligofructose One Trevis™ capsule TDS 7.5 g Raftilose™ BD |
Placebo capsules and powdered sucrose | Until death or discharge from current hospital |
Mortality: No significant difference in mortality 49% vs. 45% (p = 0.672) Length of stay: No significant difference in LOS Median LOS 15 days [IQR: 9–26] vs. 14 days [IQR: 9–29] (p = 0.913) |
Low |
| Louzada et al. [27] |
Control: 77.0 (± 1.3) Intervention: 77.2 (± 1.3) |
49 | Older persons (65–90). Pre-frail individuals registered with single health centre | Community, Brazil | RCT- 2ry analysis |
SYNBIOTIC Lactobacillus paracasei, Lactobacillus ramnosus, Lactobaciullus acidophilus, Bifidobacterium lactis (108–109 CFU of each) Fructooligosaccharide (6 g) Two daily doses (6 g + 6 g) |
Placebo (6 g + 6 g maltodextrin) | 24 weeks |
Cognition and mood: No significant effect on mood or cognition was found MMSE: median 25.0 to 25.9 vs. median 27.0 to 26.5 (p = 0.25) GDS-15: median 3.0 to 3.0 vs 2.5 to 3 (p = 0.47) |
Low |
| Neto et al. [18] |
All participants: 67.9 (± 4.5) [60–74] |
17 | Community dwelling older adults fulfilling one of Fried’s frailty criteria | Community, Brazil | RCT |
SYNBIOTIC Lactobacillus paracasei, Lactobacillus ramnosus, Lactobaciullus acidophilus, Bifidobacterium lactis (108–109 CFU of each) Fructooligosaccharide (6 g) Once daily dose diluted in water to be drank following the last meal of the day |
Placebo (maltodextrin) | 3 months |
Physical function: No significant impact on grip strength Grip strength (N): 15.0 (± 5.2) to 15.7 (± 5.3) vs. 15.9 (± 2.7) to 17.2 (± 3.9) |
Very low |
| Östlund-Lagerström et al. [19] |
Control: 72.6 (± 5.8) Intervention: 72 (± 5.6) |
249 | Free-living older adults living in their own homes | Community, Sweden | RCT |
SYNBIOTIC (although note study refers to PROBIOTIC only) Lactobacillus reuteri DSM17938 (1 × 108 CFU/stick pack), Rhamnose, galactooligosaccharide and maltodextrin To a total weight of 1 g |
Placebo (maltodextrin) |
12 weeks |
Mood: No significant differences in anxiety or depression scores HADS: − 0.32 (± 3.5) to − 0.18 (± 4.4) vs. − 0.41 (± 2.9) to 0.005 (± 5.2) (p = 0.89) PSS: − 0.41 (± 4.9) to − 1.55 (± 4.8) vs. − 0.24 (± 5.4) to − 1.12 (± 4.4) (p = 0.57) Physical function: No significant difference in scores looking at functional wellbeing EQ-5D-5L: Index: 0 (± 0.1) to 0 (± 0.1) vs − 0.01 (± 0.1) to 0 (± 0.1) (p = 0.87) VAS: 1.89 (± 10.7) to 2.88 (± 9.8) vs. 0.60 (± 11.5) to 3.47 (± 7.7) (p = 0.66) |
Moderate |
| Shimizu et al. [34] |
Control: Median age: 74 [IQR: 64–81] Intervention: Median age: 74 [IQR: 64–82] |
77 | Septic patients placed on a ventilator within 3 days of ITU admission | Intensive Care Unit, Japan | RCT |
SYNBIOTIC Yakult BL Seichoyaku 1 × 108 Bifidobacterium Breve strain Yakult/g, 1 × 108 Lactobacillus casei strain Shirota/g Galactooligosaccharides (Oligomate S-HP, Yakult Honsha) Yakult BL Seichoyaku 3 g/day + galactooligosaccharides 10 g/day |
Standard care | Until oral intake initiated |
Mortality: No significant difference in mortality seen 10.8% vs. 8.6%% (p = 0.84) |
Very low |
RCT randomised control trial, CFU colony-forming units, IQR inter-quartile range, ADLs activities of daily living, LOS length of stay, FI frailty index, HADS Hospital Anxiety and Depression Score, PSS Perceived Stress Score, VAS Visual Analogue Scale