Human microbiome and metabolic health: An overview of systematic reviews

To summarize the microbiome's role in metabolic disorders (insulin resistance, hyperglycemia, type 2 diabetes, obesity, hyperlipidemia, hypertension, nonalcoholic fatty liver disease [NAFLD], and metabolic syndrome), systematic reviews on observational or interventional studies (prebiotics/probiotics/synbiotics/transplant) were searched in MEDLINE and Embase until September 2020. The 87 selected systematic reviews included 57 meta-analyses. Methodological quality (AMSTAR2) was moderate in 62%, 12% low, and 26% critically low. Observational studies on obesity (10 reviews) reported less gut bacterial diversity with higher Fusobacterium, Lactobacillus reuteri, Bacteroides fragilis, and Staphylococcus aureus, whereas lower Methanobrevibacter, Lactobacillus plantarum, Akkermansia muciniphila, and Bifidobacterium animalis compared with nonobese. For diabetes (n = 1), the same was found for Fusobacterium and A. muciniphila, whereas higher Ruminococcus and lower Faecalibacterium, Roseburia, Bacteroides vulgatus, and several Bifidobacterium spp. For NAFLD (n = 2), lower Firmicutes, Rikenellaceae, Ruminococcaceae, whereas higher Escherichia and Lactobacillus were detected. Discriminating bacteria overlapped between metabolic disorders, those with high abundance being often involved in inflammation, whereas those with low abundance being used as probiotics. Meta-analyses (n = 54) on interventional studies reported 522 associations: 54% was statistically significant with intermediate effect size and moderate between-study heterogeneity. Meta-evidence was highest for probiotics and lowest for fecal transplant. Future avenues include better methodological quality/comparability, testing functional differences, new intervention strategies, and considerating other body habitats and kingdoms.