Table 1.
Summary of included studies: author, year, location, design, aim, participants, antibiotic, duration, molecular techniques used, and outcomes
| Author, year, location | Design and reference group |
Aim | Study participants (n), age | Antibiotic exposure (route; duration) |
Exposure to outcome interval (sampling frequency) | Molecular technique | Key findings: Richness |
Key findings: Diversity |
Key findings: Taxonomical change |
|---|---|---|---|---|---|---|---|---|---|
| 1.Bai et al.32 China |
Cross-sectional with (control group) |
Added impact of antibiotics on microbiota changes in ALL | 33 healthy children; 10 received antibiotics 1–12 years |
Cephalosporin Penicillin (oral + IV: 10 days) |
1–4 weeks (x1) |
Next generation sequencing (NGS) | n/a |
Reduction in α-diversity Shannon index, ~2.75 (antibiotics) vs ~3.25 (control), p < .05 # Simpson index ~0.15 (antibiotics) vs ~0.09 (control) p < .05 # a |
Decreased Firmicutes/Bacteroidetes ratio by approximately one third (p < .05) No difference in abundances of Actinobacteria and Proteobacteria |
| 2. Bokulich et al,33 USA |
Cohort (baseline assessment) |
Microbiome development in first 2 years of life | 43 infants, 25 received antibiotics 0–2 years |
Cephalosporin Beta-lactams Macrolides Quinolones Nitrofurantoin (route/duration not specified) |
3–139 days (x25) |
PCR, 16s RNA gene amplification |
n/a |
No change in α-diversity after antibiotic exposure for median 52 days (13–139) Reduced β–diversity: UniFrac distance, Permutational MANOVA, R2 < 0.01, p < .001 |
No effect on Bifidobacterium abundance |
| 3.Brunser et al.34 Chile |
RCT (baseline assessment) |
Impact on microbiome of prebiotic supplement following antibiotic | 130 infants, before and after antibiotics 1–2 years |
Amoxicillin (oral; 7 days) |
1–3 weeks (x3) |
FISH* and flow cytometry | n/a | n/a |
Decrease in amoxicillin-associated fecal bacteria by 30% (p < .001) Increase in amoxicillin associated E. Coli Log count 4.77 ± 0.96 (baseline) vs 5.10 ± 1.39 after treatment (p = .015) No change in total counts of Bifidobacterium and Bacteroides |
| 4.Doan et al.35 Niger |
RCT (control group) |
Effects of azithromycin on gut microbiome diversity | 80 children, 40 received antibiotics 1–5 years |
Azithromycin (oral; single dose) |
5 days (x1) |
16S rRNA sequencing | n/a |
Reduction in α-diversity: Inverse Simpson’s α-diversity decreased (5.03 95% CI 4.08–6.14) vs placebo (6.91; 5.82 − 8.21) p = .03 Shannon’s α-diversity decreased (10.60; 95% CI 8.82–12.36) vs placebo (15.42; 13.24–17.80) p = .004 No change in β-diversity Decrease in Simpson’s community level γ diversity with azithromycin (10.10 95% CI 7.80–11.40) vs placebo (17.72;13.80–20.21) (p < .001) |
n/a |
| 5.Fouhy et al.36 Ireland |
Cohort study (control group) | Effect of antibiotics on gut microbiome | 18 children, 9 received antibiotics Newborn |
Ampicillin and gentamicin (IV:2–9 days) |
4 and 8 weeks (x2) |
High throughput sequencing of 16S rRNA | n/a | No change in α-diversity (Shannon Index) at 4 weeks (3.6) vs control (3.6) (p = .575) |
Decreased Bifidobacterium (5% vs 25%; p = .013) and Lactobacillus (1% vs 4%; p < .009) in treated group at 4 weeks vs control; no difference at 8 weeks Increased Proteobacteria (44% vs 23%; p < .005) and Enterobacteriaceae (50% vs 32%; p = .006) at 8 weeks vs control. Increased Clostridium in treated infants than controls at week 8 (7% vs 2%, p < .035) |
| 6.Korpela et al.37 Finland |
Retro-spective controlled cohort study | Antibiotic induced changes in microbiota composition | 142 children, 99 received antibiotics 2–7 years |
Macrolides Penicillins (route/duration not specified) |
Variable (<6 months to 2 years) (x1-2) |
DNA extraction, 16s rRNA gene sequencing | Macrolides: Reduced richness up to 2 years (p < .05)# Penicillins: Reduced richness over 6 months# (p < .001)resolved by 12 months |
n/a | Macrolides: Exposure over 6 months Reduced Bifidobacterium (0.23-fold change p < .004) and Lactobacillus (0.12-fold change p < .004) Increased Bacteroides (2.04-fold change p < .004) and Proteobacteria (1.96-fold change (p < .02). Penicillins: Decreased Lactobacillus (0.09 fold; p < .004) with exposure in previous 12 months |
| 7.Mangin et al.38 Chile |
Cohort (baseline assessment) |
Impact of amoxicillin on fecal bifidobacteria | 31 infants, all received antibiotics 12–24 months |
Amoxicillin (oral: 7 days) |
0 days (x2) |
Total DNA extraction, PCR | n/a | n/a |
No change in total Bifidobacteria Disappearance of Bifidobacterium adolescentis species (0% vs 36.4% (p < .001) |
| 8. Oldenburg et al.39 Burkina Faso |
RCT (baseline assessment) |
Investigate effect of 3 antibiotics on microbial diversity | 124 children, 93 received antibiotics 6–59 months |
Amoxicillin Azithromycin Cotrimoxazole (oral: 5 days) |
5 days (x2) |
DNA extraction, deep gene sequencing | n/a |
Reduced α- diversity with Azithromycin: Inverse Simpsons’ index decreased (6.6 95% CI 5.5–7.8) vs baseline (8.8 95% CI 7.5–10.1) (p < .001) Shannon index decreased (11.0 95% CI 9.3–12.7) vs baseline (14.6 95% CI 13.0–16.2) (p < .001) No reduction with Amoxicillin or Cotrimoxazole |
n/a |
| 9.Parker et al.40 India |
RCT (baseline assessment) |
Assess microbiota changes following azithromycin | 114 infants, 56 received antibiotics 6–11 months |
Azithromycin (oral: 3 days) |
12 days (x2) |
16s rRNA gene sequencing, DNA extraction PCR |
Lower OTU with azithromycin: (68.1 ± 15.4) vs placebo (73.6 ± 13.7) (linear regression p = .027) c. 7% less | No significant change in α-diversity (Shannon index) azithromycin (2.6 95%CI 2.47–2.73) vs placebo (2.8 95%CI 2.8(2.67–2.93) (p = .087) |
Decreased relative abundance of Proteobacteria (mainly Escherichia) mean % ± SD: 15.9 ± 13.2 vs 10.2 ± 15.4 FDR (p < .001) and Verrucomicrobia (genus Akkermansia muciniphilia) 0.5 ± 3.1 vs 0.0 ± 0.0 FDR (p < .012) No change in Actinobacteria, Bacteroides and Firmicutes, Bifidobacterium |
| 10.Penders et al.41 The Netherlands |
Cross Sectional (control group) |
Examine contribution of external influences to gut microbiota composition | 1032 infants, 28 received antibiotics 1 month |
Mainly Amoxicillin (oral: duration not specified) |
<1 month (x1) |
DNA isolation, PCR | n/a | n/a |
Decrease with antibiotics in Bifidobacteria. Antibiotics (10.29 CFU/g log10) vs Control (10.7 CFU/g log10 (p < .01) Decrease in Bacteroides fragilis. Antibiotics (6.39 CFU/g log10) vs Control (9.31 CFU/g log10) (p < .01) No change in Lactobacilli, Escherichia coli, Clostridium difficile |
| 11. Wei et al.29 Denmark |
RCT (control group) |
Examine short- and long-term impacts of azithromycin treatment on gut microbiota in children | 72 children, 33 episodes of asthma-like symptoms received antibiotics 12–36 months |
Azithromycin (oral: 3 days) |
14 days and up to 4 years (x3) |
DNA extraction and sequencing | Decrease in richness at 14 days: Observed richness: 23% reduction (177.8 ± 56.0 vs. 230.6 ± 61.2, p < .001); no difference by mean 233 days |
Reduced α-diversity: at 14 days: Shannon index: 13% reduction; 2.96 ± 0.80 (mean ± SD) vs control 3.41 ± 0.58, p = .009) Reduced β–diversity: UniFrac distance, treatment accounted for variance (R2 = 3.8%, p = .027 (weighted) and F2 = 4.2% p < .001 (un-weighted) |
Reduction, 50-fold, in genus Bifidobacterium at 14 days (p adjusted <0.011 (FDR p < .05) Long term (13–39 months) no differences seen between azithromycin and placebo groups |
| 12.Yassour et al.42 Finland |
Cohort (control group) |
Study development of infant gut microbiome and effect of antibiotics | 39 children, 20 received antibiotics 2–36 months |
Amoxicillin Cefalexin Clarithromycin Amoxicillin and clavulanic acid Trimethoprim and sulfadiazine Azithromycin Cefaclor Penicillin G Netilmicin (oral: duration not specified) |
<1 month (x28) |
16S rRNA gene and whole genome shotgun sequencing |
n/a | Reduced microbiome strain (subspecies) diversity (diversity index 0.0003 vs 0.55 (control) (p < .001) |
Decreased abundance of species from clostridium clusters IV and XIVa (T regulatory immune cells) at aged 3 (median abundance ~9% vs ~15% control)b (p = .037) Less stable gut microbiome following antibiotic treatment (Jaccard Index P = <0.001) |
Key: # – Approximate mean values taken from a Box and Whisker plot; ꝉ – statistical significance testing, confidence intervals, or standard deviations not given; a – higher the Simpson index, the lower the diversity; b -raw data not given, approximate values taken from graph
Abbreviations: n/a- data not available. ALL: acute lymphoblastic leukemia. CFU/g = Colony forming units per gram of sample. FISH: Fluorescent in-situ hybridization. FDR – False Discovery Rate correction. OTU count: operational taxonomical unit.