Table 1.
Animal and human studies investigating the bidirectional relationship between antidepressant use and gut microbiota
| Reference | Study design | Study subjects, sample size (N), and sample characteristics | Main outcome(s) | Studied antidepressant(s), dose(s), and time of exposure (if available) | Measurement technique | Main findings |
|---|---|---|---|---|---|---|
| Preclinical studies | ||||||
| (An et al., 2020) | Cross-sectional | C57BL/6J mice, CUMS HCs + vehicle N = 5 CUMS + vehicle N = 5 CUMS + NMDEA N = 5 |
Bacterial taxonomic composition, α-diversity (Observed OTUs, Shannon, ACE, Simpson), β-diversity (unweighted UniFrac) |
NMDEA (analogue of agomelatine) 30 mg/kg/day 28 days |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) |
CUMS disrupted the richness of the gut bacterial community. NMDEA administration made the gut microbiome more similar to that of HCs. |
| (Aydin et al., 2021) | Cross-sectional | Sprague–Dawley rats Normoglycemic group (N = 12, reboxetine = 6), diabetic group (N = 12, reboxetine = 6) |
Bacterial taxonomic composition, α-diversity (observed OTUs, Shannon index, phylogenetic diversity), β-diversity (unweighted and weighted UniFrac) |
Reboxetine 8 mg/kg/day 2 weeks |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) |
Reboxetine use was associated with significant changes in specific taxa and led to an altered gut microbiome with increased inflammatory capacity. |
| (Cussotto et al., 2019) | Cross-sectional | Sprague–Dawley rats Controls N = 8 Escitalopram N = 8 Venlafaxine N = 8 Fluoxetine N = 8 Lithium N = 8 Valproate N = 8 Aripiprazole N = 8 |
Bacterial taxonomic composition, α-diversity (Shannon, Chao1), β-diversity (Bray Curtis) |
Escitalopram: 6.38 mg/kg/day venlafaxine: 20 mg/kg/day fluoxetine: 10 mg/kg/day 28 days |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) |
Microbial richness and diversity in antidepressant-administered animals did not significantly differ from those of controls. Fluoxetine administration produced a decrease in the phylum Deferribacteres and in the genera Prevotella and Succinivibrio. |
| (Dangoor et al., 2021) | Longitudinal | Wistar rats, Corticosterone-induced stress HCs N = 30 Corticosterone-induced stress N = 40 |
Microbiota composition shift pre-post treatment: (Beta diversity: weighted and unweighted UniFrac) |
Citalopram 10 mg/kg/day 2 weeks |
Amplification and sequencing of 16S rRNA genes | Both oxytocin administration and corticosterone stress induction led to significant changes in bacterial composition. No such change was observed with citalopram. |
| (Dethloff et al., 2020) | Longitudinal | DBA/2J mice, depressed mouse strain Paroxetine N = 17 vehicle N = 17 |
Bacterial taxonomic composition α-diversity (Faith’s Phylogenetic Diversity) β-diversity (unweighted and weighted UniFrac) Bile acid levels |
Paroxetine 10 mg/kg/day 2 weeks |
Amplification and sequencing of 16S rRNA genes | There were several primary and secondary bile acid level and microbiota α-diversity differences between paroxetine- and vehicle-treated mice |
| (Diviccaro et al., 2022) | Longitudinal | Sprague–Dawley rats N = N/A |
Bacterial taxonomic composition α-diversity (Evenness, Shannon, Faith, and Observed Features metrics) β-diversity (Bray Curtis, Unweighted and Weighted Unifrac) |
Paroxetine 10 mg/kg/day 2 weeks |
Amplification and sequencing of 16S rRNA genes (V4 regions) | Paroxetine did not affect α-diversity but altered β-diversity. Taxa belonging to the Firmicutes phylum were significantly altered with a significant increase in the Enterococcaceae family, belonging to the Bacilli class, and several variations related to the Clostridia class. |
| (Dong et al., 2023) | Longitudinal | Sprague–Dawley rats, CUMS HCs N = 10 HCs + fluoxetine N = 10 CUMS N = 10 CUMS + fluoxetine N = 10 |
Bacterial taxonomic composition α-diversity (Shannon, Simpson, Chao 1, ACE) β-diversity (Unweighted Unifrac) |
Fluoxetine 1 mg/kg/day 4 weeks |
Amplification and sequencing of 16S rRNA genes (V4 region) |
Fluoxetine administration did not restore CUMS induced alterations. Possible fluoxetine under dosage. |
| (Duan et al., 2021) | Longitudinal | C57BL/6 mice, CUMS HCs N = 8 CUMS+vehicle N = 7 Responders N = 7 Non-responders N = 9 |
Bacterial taxonomic composition α-diversity (Shannon, ACE, Chao1, Simpson) β-diversity (unweighted UniFrac) Serum metabolic signatures |
Escitalopram 10 mg/kg/day 4 weeks |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) |
Increased α-diversity of the gut microbiome in the escitalopram treatment group. The responder group was mainly characterized by increased levels of genus Prevotellaceae UCG-003, and the non-responder group by depleted families Ruminococcaceae and Lactobacillaceae |
| (Fung et al., 2019) | Longitudinal | C57Bl/6 mice N = 14 |
Bacterial taxonomic composition α-diversity (Observed OTUs) |
Fluoxetine 10 mg/kg/day 7 days |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
Fluoxetine administration led to a decreased abundance of Turicibacter and Clostridiaceae, with no significant differences in α-diversity. |
| (Getachew et al., 2018) | Cross-sectional | Wistar rats Saline N = 5 Ketamine N = 5 |
Bacterial taxonomic composition α-diversity (Observed species, Shannon) |
Ketamine 2.5 mg/kg/day 7 days |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) |
Ketamine reduced levels of opportunistic pathogens like Ruminococcus and Mucispirillum and elevated those of the beneficial genera Lactobacillus, Sarcina, and Turicibacter. |
| (Huang et al., 2019) | Cross-sectional | C57BL/6 mice, inflammation model of depression (LPS) HCs N = 8 LPS+ saline N = 8 LPS+ ketamine N = 8 |
Bacterial taxonomic composition α-diversity (Shannon, Simpson, Chao1) β-diversity (Bray Curtis) |
Ketamine hydrochloride 10 mg/kg single intraperitoneal administration |
Amplification and sequencing of 16S rRNA genes (V4-V5 regions) |
The phylum Actinobacteria, the class Coriobacteriia, the order Clostridiales, the family Prevotellaceae, and the genus Alloprevotella independently correlated with treatment outcomes and might be potential biomarkers for ketamine’s antidepressant efficacy. |
| (Kim et al., 2021) | Cross-sectional | C57BL/6 N mice, immobilization stress, Escherichia Coli stress HCs: N = 6 immobilization stress N = 6 immobilization stress + Buspirone orally N = 6 immobilization stress + Buspirone intraperitoneally N = 6 HCs N = 6 Escherichia Coli stress N = 6 Escherichia Coli stress + Buspirone orally N = 6 Escherichia Coli stress + Buspirone intraperitoneally N = 6 |
Bacterial taxonomic composition α-diversity (Observed OTUs) β-diversity (generalized UniFrac analysis) |
Buspirone - orally: 5 mg/kg/day, - intraperitoneally injected: 1 mg/kg/day 5 days |
Amplification and sequencing of 16S rRNA genes (V4 regions) | Buspirone partially minimized differences in gut microbiota between the stressed group and HCs (including β-diversity). |
| (Lukić et al., 2019) | Cross-sectional | BALB/c OlaHsd mice, depressed mouse strain HCs N = 9 Fluoxetine N = 11 Escitalopram N = 12 Venlafaxine N = 12 Duloxetine N = 11 Desipramine N = 12 |
Bacterial taxonomic composition α-diversity (Faith’s phylogenetic diversity, Chao1, Gini coefficient) β-diversity (unweighted and weighted UniFrac) |
Fluoxetine: 10 mg/kg/day escitalopram: 10 mg/kg/day venlafaxine: 10 mg/kg/day duloxetine: 10 mg/kg/day desipramine: 20 mg/kg/da 3 weeks |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
All the drugs except desipramine reduced the richness of the gut microbiome but did not affect its evenness, while simultaneously β-diversity of antidepressant groups was found to be higher than that of control samples. Supplementation with Ruminococcus Flavefaciens was able to abolish the antidepressant effects of duloxetine |
| (Lyte, Daniels and Schmitz-Esser, 2019) | Longitudinal | CF-1 mice HCs: N = 10 Fluoxetine N = 10 |
Bacterial taxonomic composition α-diversity (observed species, Shannon, ACE, Chao1, Simpson) |
Fluoxetine 20 mg/kg/day 29 days |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
Fluoxetine induced a significant, time-dependent decrease in the abundance of beneficial Lactobacillus johnsonii spp. |
| (McVey Neufeld et al., 2019) | Longitudinal | BALB/c mice, depressed mouse strain N = 5 |
Bacterial taxonomic composition α-diversity (Shannon, Simpson) β-diversity (Bray Curtis) |
Sertraline 6 mg/kg/day 2 weeks |
Amplification and sequencing of 16S rRNA genes (V3 regions) |
While sertraline treatment did not alter the overall microbial profile, there was a significant decrease in α-diversity over the treatment period that was not observed in controls. |
| (Qu et al., 2017) | Cross-sectional | C57BL/6 mice, social defeat stress model of depression HCs N = 6 Saline N = 6 (R)-ketamine N = 6 Lanicemine N = 6 |
Bacterial taxonomic composition β-diversity (Bray Curtis) |
(R)-ketamine 10 mg/kg single intraperitoneal injection |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
(R)-ketamine exerted its antidepressant effect by increasing decreased levels of the order Bacteroidales, Clostridiales, and family Mogibacteriaceae, and decreasing the levels of the genus Clostridium and family Ruminococcaceae. |
| (Ramsteijn et al., 2020) | Longitudinal | Wistar rats, Early life stress HCs + Vehicle N = 11 Stressed + Vehicle N = 8 HCs + Fluoxetine N = 8 Stressed + Fluoxetine N = 9 Early life stress |
Bacterial taxonomic composition α-diversity (Shannon) β-diversity (weighted UniFrac) Inferred metabolic pathways (PICRUSt) |
Fluoxetine 10 mg/kg/day during pregnancy and lactation |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
Fluoxetine treatment modulated key aspects of maternal microbial community dynamics and metabolite output, mainly in early-life stressed females. |
| (Shen et al., 2023) | Longitudinal | C57/6 mice, CUMS HCs N = 10 CUMS + vehicle N = 10 CUMS + venlafaxine N = 10 |
Bacterial taxonomic composition α-diversity (Shannon, Simpson, Chao and Ace) |
Venlafaxine 12.5 mg/kg/day 2 weeks |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
Venlafaxine made the gut microbiome more similar to that of HCs restoring CUMS alterations. |
| (Sun et al., 2019) | Cross-sectional | C57/6 mice, CUMS HCs N = 10 CUMS + vehicle N = 10 CUMS+ fluoxetine N = 10 |
Bacterial taxonomic composition α-diversity (Shannon) β-diversity (Bray Curtis) |
Fluoxetine 12 mg/kg/day 3 weeks |
Amplification and sequencing of 16S rRNA genes | Fluoxetine made the gut microbiome more similar to that of HCs |
| (Vuong et al., 2021) | Longitudinal | C57BL/6 J mice Saline N = 6 Fluoxetine N = 6 Antibiotics + saline N = 6 antibiotics + fluoxetine N = 6 |
Bacterial taxonomic composition α-diversity (Shannon) β-diversity (weighted UniFrac) |
Fluoxetine 10 mg/kg/day 8 days |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
Maternal fluoxetine treatment had no overt effect on the composition of the maternal gut microbiota. |
| (Wang et al., 2021) | Longitudinal | C57BL/6 mice, inflammation model of depression (LPS-induced) N = N/A |
Bacterial taxonomic composition α-diversity (Shannon, Simpson) β-diversity (Weighted UniFrac, Binary Jaccard) |
(S)-norketamine, (R)-norketamine 10 mg/kg single intraperitoneal injection |
Amplification and sequencing of 16S rRNA genes (V3-V5 regions) | (S)-nor-ketamine exerted potent antidepressant-like effects in lipopolysaccharide-induced mice, as it made the gut microbiota composition of stressed mice more similar to that of HCs. |
| (Yang et al., 2017) | Cross-sectional | C57BL/6 mice, social defeat stress model of depression HCs N = 6 Saline N = 6 (R)-ketamine N = 6 (S)-ketamine N = 6 |
Bacterial taxonomic composition β-diversity (Bray Curtis, Euclidean dissimilarity) |
(R)-ketamine, (S)-ketamine 10 mg/kg single intraperitoneal administration |
Amplification and sequencing of 16S rRNA genes (V4 regions) |
Gut microbiota partially mediated the antidepressant actions of (R)-ketamine, mainly through the elevation of decreased levels of the Mollicutes class and Butyricimonas genus. |
| (Zhang et al., 2021) | Longitudinal | Sprague–Dawley rats, CUMS HCs N = 12 CUMS N = 36 (Fluoxetine N = 7, Amitriptyline N = 6) |
Bacterial taxonomic composition α-diversity (Shannon, Simpson, Chao1) β-diversity (unweighted and weighted UniFrac) Inferred metabolic pathways |
Fluoxetine: 12 mg/kg/day amitriptyline: 25 mg/kg/day 8 days |
Amplification and sequencing of 16S rRNA genes | Administration of amitriptyline and fluoxetine reversed part of the gut microbiota profile and functions altered by CUMS. |
| Clinical studies | ||||||
| (Bharwani et al., 2020) | Longitudinal | MDD (N = 15) F/M:12/3 Mean age: 36.9 ± 12.9 |
Bacterial taxonomic composition α-diversity (Faith’s phylogenetic diversity) β-diversity (Bray-Curtis) |
Citalopram, escitalopram 6 months |
Amplification and sequencing of 16S rRNA genes | Remitters showed higher baseline α-diversity, which remains evident following 6 months of treatment. Bacteria from the Clostridiales order were elevated only in remitters after 6 months of treatment vs baseline. 35 OTUs were significantly different between remitters and nonremitters at 3-month follow-up, and 42 OTUs were different at 6-month follow-up. |
| (Dong et al., 2022) | Longitudinal | MDD N = 63 HCs N = 30 F/M: 63/30 Mean age: 28.34 (MDD), 29.23 (HCs) |
Bacterial taxonomic composition α-diversity (Chao1, Shannon, Simpson) β-diversity (Bray Curtis) Metabolic pathways (GC-MS) |
Citalopram, escitalopram, paroxetine, venlafaxine 42.33 ± 11.88 mg/day (Fluoxetine-equivalent dose) 8 weeks |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) | The abundance of the phylum Actinobacteria, families Christensenellaceae and Eggerthellaceae, genera Adlercreutzia, and Christensenellaceae R-7 group was significantly lower at baseline in responders than in non-responders. |
| (Falony et al., 2016) | Cross-sectional | General population N = 2241 Mean age: 50,9 ± 13,6 |
Bacterial taxonomic composition α-diversity (observed genera, Pielou’s index, Fisher’s α) β-diversity (Bray-Curtis, Jensen-Shannon) |
Venlafaxine | Amplification and sequencing of 16S rRNA genes (V4 regions) | There was a significant correlation between venlafaxine use and gut microbiome composition, with an increase in the genus Clostridium IV. |
| (Fontana et al., 2020) | Cross-sectional | MDD HCs N = 20 N = 8 treatment resistant (TR) N = 19 responders (R) N = 7 untreated (U) F/M:31/23 Mean age: 58.8(TR), 53.7 (R), 57.0 (U), 37.7 (HCs) |
Bacterial taxonomic composition | SSRIs/SNRIs/TCAs/ Serotonin modulator TR = 44.0 months (median) R = 24.0 months (median) |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) | The gut microbiota of TR patients significantly differed from responders for the families of Paenibacillaceae and Flavobacteriaceaea, for the genus Fenollaria, and the species Flintibacter butyricus, Christensenella timonensis, and Eisenbergiella massiliensis, among others. The phyla Proteobacteria, Tenericutes and the family Peptostreptococcaceae were more abundant in TR patients, whereas the phylum Actinobacteria was enriched in responders. |
| (Gao et al., 2023) | Cross-sectional, Longitudinal | MDD N = 62 N = 25 treatment resistant (TR) N = 37 responders (R) HCs N = 41 F/M: 53/41 Mean age: 24.4 (TR), 22.67 (R), 23.34 (HCs) |
Bacterial taxonomic composition | SSRIs 8 weeks |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) | The relative abundance of Blautia, Coprococcus, and Bifidobacterium was elevated in the responsive patients with MDD |
| (Imhann et al., 2016) | Cross-sectional | General population N = 1174 F/M: 688/486 Mean age: 48,36 ± 13,58 |
Bacterial taxonomic composition α-diversity (observed species, Shannon) β-diversity (Bray-Curtis) |
SSRIs, SNRIs, Mirtazapine, Tricyclics | Amplification and sequencing of 16S rRNA genes (V4 regions) | The increase of two taxa (belonging to Bacteroidetes phylum) was significantly associated with antidepressant use, but not after correction for PPI use. |
| (Jackson et al., 2018) | Cross-sectional | General population N = 2737 F/M: 2435/302 Mean age: 60 ± 12 |
Bacterial taxonomic composition α-diversity (Shannon index, phylogenetic diversity, raw OTUs counts) β-diversity (weighted and unweighted UniFrac) |
SSRIs, TCAs | Amplification and sequencing of 16S rRNA genes (V4 regions) | Significant associations between antidepressant medications and gut microbiota markers were found. SSRIs were negatively associated with Turicibacteraceae abundance. |
| (Jiang et al., 2015) | Cross-sectional | MDD N = 29 Active-MDD N = 17 responders HCs N = 30 F/M: 34/42 Mean age: 25,3 (A-MDD), 27,1 (R), 26,8 (HCs) |
Bacterial taxonomic composition α-diversity (observed OTUs, Chao1, ACE, Shannon, Simpson) β-diversity (unweighted UniFrac) Serum cytokines |
SSRIs, SNRIs 4 weeks |
Amplification and sequencing of 16S rRNA genes (V1-V3 regions) | Increased α-diversity was found in the active-MDD vs. the HC group but not in the responders vs. the HC group. Bacteroidetes, Proteobacteria, and Actinobacteria strongly increased in level, whereas Firmicutes were significantly reduced in the A-MDD and R-MDD groups compared to the HC group. |
| (Lin et al., 2024) | Cross-sectional | Mood Disorders N = 271 F/M: 172/89 Mean age: 44.31 (SSRI/SNRI), 43.25 (Other AD), 43.41 (No AD) |
Bacterial taxonomic composition α-diversity (Faith’s, Shannon) β-diversity (unweighted UniFrac) |
SSRIs, SNRIs | Amplification and sequencing of 16S rRNA genes (V3-V4 regions) | SSRIs and SNRIs showed reduced alpha diversity. |
| (Kurokawa et al., 2021) | Cross-sectional, Longitudinal | MDD N = 16 (non-responders) N = 11 (responders) N = 6 (stable remitters) F/M:17/16 Mean age: 49.1 (responders), 53 (non-responders), 56.3 (stable remitters) |
Bacterial taxonomic composition α-diversity (Chao 1, Shannon, Phylogenetic diversity) β-diversity (unweighted and weighted UniFrac) Inferred Metabolic pathways (PICRUSt) |
Mixed antidepressants | Amplification and sequencing of 16S rRNA genes (V1-V2 regions) | α-diversity was lower in non-responders compared to responders during the treatment course, while no significant differences in baseline α-diversity indices were detected. Non-responders presented increased estimated glutamate synthesis microbiota functions compared to responders and stable remitters. |
| (Lee et al., 2022) | Pilot randomized controlled trial | MDD Placebo N = 10 Levomilnacipram N = 7 F/M:7/10 Mean age: 73 (placebo), 70 (levomilnacipram) |
Response to antidepressant treatment (Depressive symptom severity at baseline and follow-up assessed using the HDRS24) | Levomilnacipram 120 mg/day 12 weeks |
Amplification and sequencing of 16S rRNA genes (V4 regions) | Baseline enrichment of Faecalibacterium, Agathobacter, and Roseburia was associated with remission. |
| (Liśkiewicz et al., 2019) | Longitudinal | MDD N = 17 F/M: 9/8 Mean age: 42.5 ± 13.9 years |
Bacterial taxonomic composition α-diversity (Observed OTUs, Chao1, Shannon, inverse Simpson) β-diversity (Bray-Curtis) |
Escitalopram 5–20 mg/day 6 weeks |
Amplification and sequencing of 16S rRNA genes (V4 regions) | Escitalopram treatment led to increased α-diversity, although no significant differences in taxa abundance were found at different time points. |
| (Liśkiewicz et al., 2021) | Longitudinal | MDD N = 16 F/M: 8/8 Mean age: 42.9 ± 14.3 years |
Severity of depressive symptoms (HDRS24), Therapeutic response (early improvement, treatment response, remission from depression, dHDRS24) |
Escitalopram 5–20 mg/day 6 weeks |
Amplification and sequencing of 16S rRNA genes (V4 regions) | The baseline abundance of different taxa was associated with the severity of depressive symptoms, but no changes in the microbiota composition were detected after clinical improvement. |
| (Madan et al., 2020) | Longitudinal | N = 111 (SUD, MDD, Bipolar Spectrum, Anxiety Spectrum, Psychotic Spectrum, Personality Disorders) F/M: 60/51 Mean age: 35.7 ± 13.8 |
Suicide, trauma, depression and anxiety severity, treatment response | Mixed antidepressants | Amplification and sequencing of 16S rRNA genes (V4 regions), whole genome shotgun sequencing | Gut microbiota richness and α-diversity early in the course of hospitalization were significant predictors of depression remission at discharge. |
| (Minichino et al., 2023) | Meta-analysis of longitudinal and cross-sectional studies (only longitudinal studies were meta-analyzed for antidepressants) | 19 studies included (8 on subjects treated with antidepressants) | Bacterial taxonomic composition α-diversity β-diversity Inferred metabolic pathways Association between gut microbiome and measures of treatment response and tolerability |
Escitalopram, vortioxetine, mixed antidepressants |
- | Significant changes were found in β-diversity but not α-diversity metrics following treatment with antidepressants. Antidepressant treatment was consistently associated with an increased abundance of the Christensenellaceae family. |
| (Rogers and Aronoff, 2016) | Cross-sectional | Community residents N = 1135 F/M: 103/48 Mean age: 52 |
Bacterial taxonomic composition α-diversity (Shannon, inverse Simpson) |
Mixed antidepressants (cited citalopram) | Amplification and sequencing of 16S rRNA genes (V3-V5 regions) | Citalopram was significantly associated with an increased abundance of a specific OTU of the Enterobacteriaceae family. Moreover, adding antidepressants to NSAIDs led to changes in gut microbiota composition. |
| (Shen et al., 2021) | Longitudinal | MDD N = 30 HCs N = 30 F/M:28/32 Mean age: 44.83 (MDD), 43.97(HCs) |
Bacterial taxonomic composition α-diversity (ACE, Chao1, Shannon, Simpson) β-diversity (Binary Jaccard) Inferred Metabolic pathways (PICRUSt) |
Escitalopram 16.33 ± 3.46 mg/day 4–6 weeks |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) | The intestinal flora of depressed patients tended back to normal under escitalopram treatment. |
| (Stanislawski et al., 2021) | Cross-sectional | N = 331, BD1/2 (17%), anxiety disorders (38%), PTSD (49.8%) F/M: 56/275 Mean age: 47.6 |
Bacterial taxonomic composition α-diversity (Observed species, Shannon diversity, and Faith’s phylogenetic diversity) β-diversity (unweighted and weighted UniFrac, Jaccard, and Bray-Curtis) Inferred Metabolic pathways (PICRUSt) |
Atypical antidepressants (trazodone, bupropion), SSRIs, or SNRIs | Amplification and sequencing of 16S rRNA genes | SNRI use significantly correlated with unweighted UniFrac, implying that it is associated with significant shifts in some taxa. Furthermore, the serotonin antagonist and reuptake inhibitor (SARI) trazodone showed an inverse relationship with α-diversity. |
| (Thapa et al., 2021) | Cross-sectional | MDD N = 110 HCs N = 27 PCs N = 23 F/M:92/98 Mean age: 19.5 (MDD), 26.3 (HCs), 19.1 (PCs) |
Bacterial taxonomic composition α-diversity (Observed OTUs, Chao1, ACE, Phylogenetic diversity, Shannon) β-diversity (unweighted and weighted UniFrac, Bray Curtis) |
SSRIs | Amplification and sequencing of 16S rRNA genes (V4 regions) |
Using SSRIs was not associated with differential bacterial composition or diversity. |
| (Ticinesi et al., 2017) | Cross-sectional | N = 76, Elderly patients (chronic comorbidities) HCs N = 25 F/M: 37/39 Mean age: 83.3 ± 7.5 |
Bacterial taxonomic composition α-diversity (Chao1, Shannon) β-diversity (unweighted UniFrac) |
Mixed antidepressants | Amplification and sequencing of 16S rRNA genes (V3 regions) | Antidepressant use was positively correlated with several taxa (Asteroleplasma, Helicobacter, Marinilactibacillus, unclassified members of the Bacilli class, and Succinivibrionacae family) |
| (Tomizawa et al., 2021) | Longitudinal | MDD N = 24 Persistent depressive disorder N = 8 Panic disorder N = 1 General anxiety disorder N = 6 Social anxiety disorder N = 1 F/M:23/17 Mean age: 54.4 ± 19.00 |
α-diversity (Chao1, Shannon, Faith’s phylogenetic diversity) β-diversity (unweighted and weighted UniFrac) Inferred Metabolic pathways (PICRUSt, GBM) |
Amitriptyline, amoxapine, sertraline, paroxetine, escitalopram, duloxetine, venlafaxine, milnacipran, mirtazapine 142 ± 364.50 days |
Amplification and sequencing of 16S rRNA genes (V1-V2 regions) | There was no significant diversity or compositional differences between individuals taking or not antidepressants. Antidepressant users showed increased gamma-aminobutyric acid (GABA) III synthesis and GABA degradation. |
| (Vich Vila et al., 2020) | Meta-analysis | N = 1124 General population (GP) N = 305 IBS N = 454 IBD F/M: 1120/763 Mean age: 44.8 (GP), 45.4 (IBS), 42.8 (IBD) |
Bacterial taxonomic composition α-diversity (Shannon) β-diversity (Bray-Curtis) Inferred metabolic pathways (PICRUSt) |
SSRIs (paroxetine, fluoxetine, sertraline, citalopram, escitalopram, fluvoxamine), TCAs (amitriptyline, maprotiline, clomipramine), other antidepressants (venlafaxine, duloxetine, mirtazapine, bupropion, trazodone) |
Amplification and sequencing of 16S rRNA genes | The abundance of Streptococcus salivarius was increased in SSRI users. An increased abundance of Eubacterium ramulus was specific to participants using SSRIs. The use of TCAs was associated with an increased abundance of Clostridium leptum. The pathway involved in peptidoglycan maturation was decreased in the multi-drug meta-analysis of SSRI users compared to non-users. |
| (Vogt et al., 2017) | Cross-sectional | Alzheimer’s disease N = 25 HCs N = 25 F/M: 35/15 Mean age: 71.3 (Alzheimer’s disease), 69.3 (HCs) |
Bacterial taxonomic composition α-diversity (ACE, Chao1, Inverse Simpson, Shannon Index) β-diversity (Bray-Curtis, and weighted and unweighted UniFrac) |
SSRIs | Amplification and sequencing of 16S rRNA genes (V4 regions) | There was no correlation between microbiota composition and diversity and antidepressant use. |
| (Wang et al., 2023) | Longitudinal | MDD N = 110 HCs N = 166 F/M:167/109 Age range: 18–65 |
Bacterial taxonomic composition α-diversity (Chao1, Inverse Simpson, Shannon Index) β-diversity (Bray-Curtis) Inferred metabolic pathways Fecal and plasma metabolite profiling |
Escitalopram 10–20 mg/day 12 weeks |
Shotgun metagenomic sequencing | The treatment with escitalopram did not effectively shift the gut microbiome of MDD patients towards a healthier or more similar state to that of HCs. |
| (Ye et al., 2021) | Longitudinal | MDD N = 26 HCs N = 28 F/M:42/12 Mean age: 26.04 (MDD), 26.04 (HCs) |
Bacterial taxonomic composition α-diversity (Observed OTUs, Chao1, Shannon) β-diversity (unweighted UniFrac) |
Vortioxetine 10 mg/day 4–8 weeks |
Amplification and sequencing of 16S rRNA genes (V3-V4 regions) | Vortioxetine ameliorated depressive symptoms by promoting the reconstruction of the gut microbiota. |
| (Zhang et al., 2019) | Pilot clinical trial | IBS + depression N = 6 Duloxetine-treated N = 9 Bifico-treated |
Bacterial taxonomic composition α-diversity (Shannon) Serum cytokines Fecal short-chain fatty acids |
Duloxetine 30–60 mg/day 8 weeks |
Amplification and sequencing of 16S rRNA genes (V1-V3 regions) | The overall gut microbiota profile shifted after treatment with duloxetine. While the α-diversity of gut microbiota did not significantly change after treatment, the abundance of the genera Fecalibacterium, Lachnospiraceae incertae sedis, Escherichia/Shigella, and Sutterella showed a tendency to increase, and Erysipelotrichaceae incertae sedis tended to decrease. |
| (Zhernakova et al., 2016) | Cross-sectional | General population N = 1135 F/M: 661/474 Mean age: 45.04 ± 13.60 |
Bacterial taxonomic composition α-diversity (Shannon, CoG richness, Gene richness) β-diversity (Bray-Curtis) Inferred metabolic pathways (PICRUSt) |
TCAs, SSRIs, other antidepressants | Amplification and sequencing of 16S rRNA genes (V4 regions) | SSRIs show a positive association with Shannon’s diversity index, while other antidepressants showed a negative association. TCA use was significantly associated with an increase in two species from the genus Bacteroides and Coprococcus and a decrease in a species from the genus Eubacterium. There was a significant positive correlation of SSRIs with the lyxose degradation pathway and a negative correlation of other antidepressants with the methylerythritol phosphate pathway. |
ACE, abundance-based coverage estimator; BD1, bipolar disorder type 1; CUMS, chronic unpredictable mild stress; dHDRS24, 24-item Hamilton Depression Rating Scale score change from baseline to endpoint; F/M, female/male ratio; HCs, healthy controls; IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; MDD, major depressive disorder; NMDEA, N-(2-(7-methoxy-3,4-dihydroisoquinolin-1-yl)ethyl)acetamide hydrochloride; OTUs, Operational Taxonomic Units; PICRUSt, Phylogenetic Investigation of Communities by Reconstruction of Unobserved States; PCs, psychiatric controls; PPIs, proton pump inhibitors; PTSD, post-traumatic stress disorder; SNRIs; serotonin-norepinephrine reuptake inhibitors; SSRIs, selective serotonin reuptake inhibitors; TCAs, tricyclic antidepressants; TR, treatment-resistant.