Table 1. Characteristics of the included four studies.

Study	Region	Study participants	Number of participants [male, female]	Age, median [range] or mean ± SEM	Intervention	Reported outcome of interest	Primary outcome of the original study
Frankwich [2012] (13)	The USA	・BMI >30 kg/m2	24 [20, 4]	PLA 54.5±1.7	Placebo vs. DOX 200 mg/day ×12 weeks	HOMA-IR, body weight, rate of diarrhea, FPG, and all adverse events	C-reactive protein and myeloperoxidase levels
		・7.5%< HbA1c <10%		DOX 55.3±1.9
Vizuete [2012] (26)	The USA	・BMI >30 kg/m2	66 [18, 48]	All 45 [18–62]	Placebo vs. Rifaximin 1,100 mg/day ×20 days	No outcome of interest was reported	Weight loss
Reijnders [2016] (15)	Netherlands	・BMI 25-35 kg/m2	57 [57, 0]	PLA 60.9±1.7 AMOX 55.7±1.5 VANCO 60.6±1.5	Placebo vs. AMOX 1,500 mg/day ×7 days vs. VANCO 1,500 mg/day ×7 days	HOMA-IR, rate of diarrhea, FPG, GLP-1, fecal SCFAs, and all adverse events	Insulin sensitivity measured by using the hyperinsulinemic-euglycemic clamp technique
		・HOMA-IR >2.2
		・Two-h PG during 75 g OGTT 7.8-11.1 mmol/l and/or fasting PG ≥5.6 mmol/L
		・Body weight stable for at least three months (±3 kg)
Balliu [2017] (28)	The USA	・Overweight and obese individuals	18 (not reported]	Not reported	Placebo vs. DOX ×12 weeks	No outcome of interest was reported	The change in insulin sensitivity measured by both HOMA and the area under the curve for c-peptide during an OGTT

AMOX, amoxicillin; BMI, body mass index; DOX, doxycycline, GLP, glucagon-like peptide; HOMA-IR, homeostasis model assessment of insulin resistance; OGTT, oral glucose tolerance test; PG, plasma glucose; RCT, randomized controlled trial; SD, standard deviation; SEM, standard error of mean; VANCO, vancomycin.