Table 1.

Summary of results.

Design, N	Results in Diversity between SCZ/HC	Results in Clinical Characteristics	Major Limitations	Reference
Case– control with prospective cohort component at 12 months FEP (n = 28) HC (n = 16)	None reported	Lachnospiraceae, Bacteroides spp. and Lactobacillus correlated with increased psychotic symptoms. Lachnospiraceae, Ruminococcaceae, spp. associated with negative symptoms. Lactobacillus correlated with increased positive symptoms. Decreased GAF correlated to Ruminococcaceae, Bacteroides, spp. Lactobacillus. Microbiota clustering at intake correlated with remission at 12 months follow-up.	Small sample size, no alpha or beta diversity reported, remission model only attempted to correlate 5 bacterial families, lack of detailed dietary information	Schwarz et al., 2018
Case–control, cross-sectional SCZ (n = 25) HC (n = 25)	Alpha: no difference Beta: significant difference	Ruminococcaceae abundance correlated with decreased negative symptoms, and Bacteroides with worse depressive symptoms. Increased genus Coprococcus associated with increased CHD risk. Phylum Cyanobacteria correlated to later disease onset, without relation to disease duration. Self-reported mental well-being correlated with phylum Verrucomicrobia.	Small sample size, no causality established, not AP naive	Nguyen et al., 2019
Cross-sectional (also included animal component not reviewed here) SCZ (n = 63) HC (n = 69)	Alpha: SCZ lower alpha diversity than HC Beta: significant difference	Symptom severity correlated positively with Bacteroidaceae, Streptococcaceae, Lachnospiracea and negatively with Veillonellaceae.	Within humans, no temporal relationship, not AP naive	Zheng et al., 2019
Case–control SCZ (n = 82) HC (n = 80)	Alpha: no difference Beta: significant difference	Succinivibrio correlated with overall symptom severity as well as the general psychopathology. Corynebacterium negatively correlated to the severity of negative symptoms.	Not AP naïve, all SCZ group inpatients, but not HC, no causality established	Li et al., 2020
Case–control, cross- sectional SCZ with violence (n = 26) SCZ w/o violence (n = 16)	Alpha: no difference Beta: no difference	Violent features were correlated to an increased abundance of (p_Bacteroidetes, c_Bacteroidia, o_Bacteroidales, f_Prevotellaceae, s_Bacteroides_uniformis), and decreased abundance of (p_Actinobacteria, c_unidentified_Actinobacteria, o_Bifidobacteriales, f_Enterococcaceae, f_Veillonellaceae, f_Bifidobacteriaceae, g_Enterococcus, g_Candidatus_Saccharimonas, g_Bifidobacterium, and s_Bifidobacterium_pseudocatenulatum).	SCZ not AP naïve, small sample size, no causality established, only history of violence assessed, lack of diet information	Chen et al., 2021
Case–control, cross- sectional SCZ (n = 38), incl. 18 TR, treatment resistant, and 20 R, responsive. HC (n = 20)	Alpha: no difference in SCZ vs. HC Beta: significant difference No diversity measures for TR vs. R.	Treatment resistance associated with increased phyla Candidatus Saccharibacteria, and Tenericutes Genera Actynomyces and Porphyromonas. Absent in TRS but present in R were families Flavobacteriaceaea and Enterococcaceae, and species Flintibacter butyricus.	Small sample size, no causality established, not AP naïve, lacking longer-term dietary information	Manchia et al., 2021
Case–control, cross- sectional Acute (n = 42) Remission (n = 40) HC (n = 44)	Alpha: no difference between 3 groups Beta: acute group distinct from control and remission groups	Haemophilus positively correlated with negative psychiatric symptoms, Corprococcus negatively correlated with negative symptoms, abundance of Haemophilus positively correlated to excitement, cognition, and depression.	Relatively small sample size, no causality established, no information about diet collected, SCZ hospitalized	Zhu et al., 2021