Serum homocysteine showed potential association with cognition and abnormal gut microbiome in major depressive disorder

Abstract

BACKGROUND

Cognitive impairment is one of the common clinical manifestations of depression, causing negative distress to patients. Elevated homocysteine (Hcy) concentrations and gut microbiome dysfunction may be observed in patients with depression.

AIM

To investigate the relationship between Hcy, microbiome, and cognition in depressive patients.

METHODS

We recruited 67 patients with major depressive disorder (MDD) (MDD group) and 94 healthy controls (HCs) individuals (HCs group). Serum Hcy levels were determined using the enzyme circulation method. 16s rRNA sequencing was used to classify and identify the fecal bacteria. 17 Hamilton depression rating scale and MATRICS consensus cognitive battery were used to evaluate mood states and cognition in patients with MDD. Correlation analysis was performed to explore the correlation between fecal flora, Hcy, and depressive cognitive function.

RESULTS

Elevated serum levels of Hcy were seen in patients with MDD compared to healthy individuals. Patients with MDD indicated significant decreases in cognitive scores (P < 0.001) in six modules: Speed of processing, working memory, visual learning, reasoning and problem-solving, social cognition, and total scores. Hcy levels showed a negative correlation with processing speed, social cognition, and total MDD scores (P < 0.05). Hcy was also significantly negatively correlated with Alistipes, Ruminococcae, Tenericides, and Porphyromonas (P < 0.05).

CONCLUSION

Our results highlight that Hcy was correlated with cognition and gut microbiome in MDD. This interaction may be related to the physiological and pathological mechanisms underlying cognitive deficits in depression.

Key Words: Homocysteine; Microbiome; Intestinal flora; Gut microbiota; Gut–brain axis; Major depressive disorder; Cognitive function; Cognitive impairment

Core Tip: Cognitive impairment is common in patients with depression, who often exhibit elevated levels of homocysteine (Hcy) and abnormalities in gut microbiome. However, to date, no research has systematically investigated the relationship between Hcy, the microbiome, and cognition in depressive patients. Our study integrates results from functional predictive analysis, microbiome data, and clinical scale assessments, revealing a close association between Hcy levels, cognitive function, and gut microbiome, in major depressive disorder. These findings may contribute to elucidating the physiological and pathological mechanisms underlying cognitive deficits in depression, further supporting the brain-gut axis theory, and providing evidence for gut- microbiome-based therapeutic strategies.

INTRODUCTION

Major depressive disorder (MDD) is a common clinical mental disorder that has a negative impact on life expectancy and daily function. From the clinical perspective, it is characterized by three dimensions: Affective, somatic, and cognitive symptoms[1]. Cognitive impairment often manifests as impaired executive function, attention, memory, and information processing speed in patients with MDD[2,3]. A recent large-sample randomized clinical trial revealed that there might be a rare subtype of MDD with cognitive control deficits as its primary manifestation[4]. A previous study indicated that 44% of patients with MDD who had achieved complete or partial remission of emotional symptoms still exhibited cognitive dysfunction[5]. Accumulating evidence confirms that cognitive impairment is recognized as a residual symptom throughout the course of a depressive episode and lack of effective antidepressant treatment[6-8]. However, due to the complexity of cognitive function and the diversity of cognitive domains, there is no clear and consistent conclusion on the pathogenesis of cognition. Cognition in MDD requires further research.

Homocysteine (Hcy) plays an important role in the development of depression as a methionine intermediate[9]. Elevated Hcy exerts neurotoxic effects by modulating N-methyl-D-aspartate and N-glutamate receptors, thereby affecting psychiatric symptoms[9,10]. Increased Hcy level is related to a decrease in S-adenosylmethanethionine (SAM), which in turn results in an imbalance in the metabolism of monoamine transmitters[11]. SAM as a methyl donor to facilitate essential methylation reactions, plays a crucial role in the metabolism and function of monoamine neurotransmitters, such as dopamine, serotonin, and norepinephrine (NE). Furthermore, these monoamine transmitters have been reported to participate in the pathogenesis of depression. Recent studies have reported that exogenous administration of SAM can enhance the activation of 5-hydroxytryptamine (5-HT) 1 receptors and promote 5-HT signaling[12]. Additionally, elevated levels of Hcy may create a hypermethylated environment that leads to methylation variations in the promoter region of the catechol-O-methyltransferase gene, thereby affecting gene expression and interfering with the normal synthesis and degradation of monoamine neurotransmitters[13]. Moreover, research has shown that a decrease in SAM concentration may reduce the expression of the intestinal monoamine oxidase A gene, which is associated with the synthesis of serotonin and dopamine[14]. These intricate mechanisms collectively contribute to the complex relationship between Hcy and emotional state.

A systematic review also summarized that increased levels of Hcy showed a relationship with cognitive function impairment[15]. Extensive research about the relationship between Hcy and cognition focused on other psychiatric diseases including schizophrenia (SCZ), bipolar disorder, Alzheimer’s disease, etc.[16-18], only 2 researches focused on late-life depression. The co-existence of high levels of Hcy with depressive symptoms usually indicates further impairment of cognitive function in late-life depression, the above mechanism may be related to cerebral white matter abnormalities[19-22]. However, no previous study explored the association in non-geriatric depression, and the exact mechanism by which this occurs is unclear.

Recent data suggest that imbalances in the gut microbiome are relevant to the development of cognitive deficits[23], which may be related to the regulation of the microbiota-gut-brain axis. Animal studies have confirmed that probiotics have a significant improvement in cognitive behavior as well as brain function[24]. Clinical intervention experiments have confirmed that certain strains have a certain degree of improvement in cognitive function in normal people or in Alzheimer’s patients[25]. These studies suggest that changes in intestinal microbiota composition are intimately associated with the occurrence and development of cognitive impairment. Surprisingly, very little attention has been paid to the association between Hcy and gut microbiota. Recent research such as that conducted by Li et al[26] reported that alteration of gut microbiome induced homocysteinemia in mice. In turn, an increase in Hcy also disrupts the intestinal barrier[27], in which changes in the concentration of certain metabolites may be associated with the pathogenesis of MDD[28]. Similarly killing or inhibiting the growth of the intestinal microbiome with antibiotics can reduce the concentration of Hcy in the plasma of mice[26]. Most anthropological research focuses on circulatory and metabolic diseases rather than psychiatric diseases. Probiotics can reduce the level of Hcy[29], while Clostridium cluster IV and Butyricimonas raise the level of Hcy[30]. Therefore, the connection between Hcy and intestinal flora needs to be further investigated.

Dysbiosis operates in the development of depression, however, the specific pathways by which gut microbes interact with the brain are not fully understood. In the field of psychiatry, only one existing work provided important information on the association between cognition and flora in SCZ patients, data about patients with MDD is limited[31]. Therefore, it is necessary here to explore whether there is an association between cognitive function, serum Hcy levels, and gut flora in depressed patients.

MATERIALS AND METHODS

Participants

From December 2019 to June 2023, sixty-seven patients (male/female = 30/37) with MDD were recruited from Beijing Huilongguan Hospital, using the International Classification of Diseases 10^th Revision (ICD-10) for determining patients diagnoses. Ninety-four healthy controls (HCs), (male/female = 32/64) were recruited through advertisements from the nearby community. The MDD and HC groups need to be informed of the purpose of the study and the study process and sign a paper informed consent form before the trial begins. This study was approved by the Ethics Committee of the Beijing Huilongguan Hospital (No. 2019-43). The inclusion criteria for patients with MDD were as follows: (1) Age 18 to 55 years; (2) Diagnosis of MDD according to the ICD-10 (F32.1, F32.2, F33.1, F33.2); (3) Hamilton depression rating scale (HAMD-17) score > 17; and (4) Drug-naive or had not received any treatment for mental illness for at least one week before the study. The exclusion criteria for all patients and HCs were as follows: (1) Prior medical history of severe head injuries, drug addiction or misuse, intellectual handicap, central nervous system disorders, or other serious illnesses; (2) Use of probiotic, symbiotics or antibiotics 30 days prior enrollment; (3) Use of vitamin B6, B12, folic acid, and other drugs can affect Hcy metabolism 90 days prior enrollment; (4) History of gastrointestinal surgeries or serious congenital anomalies; and (5) Participation in other research projects within 3 months. At least two psychiatrists screened all the participants. Sex, age, educational level, and body mass index (BMI) were also recorded.

Measures of clinical symptoms and cognitive function

The clinical evaluation of this study was performed by two trained clinicians, we used the HAMD-17 to assess the severity of depression. It is easy to quantify depressive symptoms and is highly standardized, which makes it a commonly used tool in clinical assessments of depressive states. Generally, we believe that the overall score reflects the severity of the disease, and the HAMD-17 score > 17 indicates major depressive symptoms[32,33]. We chose the MATRICS consensus cognitive battery (MCCB) to assess cognitive function. Its Chinese version includes seven dimensions: (1) Speed of processing (SOP); (2) Attention/vigilance (CPT); (3) Working memory (WM); (4) Verbal learning (HVLT); (5) Visual learning (BVMT); (6) Reasoning and problem-solving (MAZES); and (7) Social cognition (EIT). The MCCB was initially used to assess cognitive function in SCZ. However, its assessment is no longer limited to SCZ. It is increasingly used in other mental disorders, such as autism and affective disorders[34,35], especially in patients with MDD. The use of the MCCB in depression has been shown to obtain consistent and reliable results[36].

Collection/handling of blood and fecal samples

A researcher trained in standardized laboratory operations collected and isolated our samples. Approximately 5 mL of intravenous blood was collected from 6:00 to 8:00 after fasting for at least 8 hours, starting at 22:00. The previous day. Blood samples were transferred to the Laboratory Department of Beijing Huilongguan Hospital for serum isolation and assessment immediately after acquisition. Hcy levels were measured using an enzymatic cycling assay and enzyme-linked immunosorbent assay kits. The reagents used to detect Hcy concentrations were procured from Beijing Leadman Biochemistry Co., Ltd. Measurement was performed using a Beckman Coulter AU5800. For stool samples, each participant collected ≥ 1 g of fresh stool using a special fecal collection kit (GeWei Bio-Tech Shanghai Co., Ltd) from their excreta and placed in a sterile stool collection bin, then immediately frozen at -80 °C until sent for further testing. DNA was extracted from fecal samples using the QIAamp DNA stool mini kit (QIAGEN, Hilden, Germany). The extracted genomic DNA was then subjected to 1.2% agarose gel electrophoresis.

DNA sequencing and microbiome analysis

The isolated fecal DNA was used as the amplification template, and the sequence of hypervariable region V4-V5 was selected as the amplification sequence with the next-generation sequencing. The selected polymerase chain reaction amplification primers were 515F (5’-GTGCCAGCMGCCGCGGTAA-3’) and 926R (5’-CCGTCAATTCMTTTGAGTTT-3’), and the amplification was carried out by a two-step amplification method. After selecting the 16S region of the bacterial fragments, specific labels (adapters, sequencing primers, and barcodes) were added at both ends of the target fragment for sequencing on the Illumina platform. Finally, a library was constructed using Microbial Biologics. After selecting the 16S region of the bacteria, the specific tags (junction, sequencing primer, and barcode) required for sequencing on the Illumina platform were added to both ends of the target fragments to complete library construction. Sequencing was completed using the Novaseq 6000 SP 500 cycle reagent kit (Illumina, United States) at TinyGene Bio-Tech Co., Ltd. (Shanghai, China).

For raw FASTQ data obtained from sequencing, Trimmomatic (version 0.35) was used to filter for quality. Based on the Silva 128 database, we used UPARSE software (usearch version V8.1.1756, http://drive5.com/uparse/) and chose the internationally accepted 97%[23,37,38] similarity level to cluster operational taxonomic units (OTUs). The species and OTU abundances were determined at different taxonomic levels. Community structure was statistically analyzed at the taxonomic levels of phylum, order, family, genus, and species, with a confidence threshold of 70%. Alpha diversity indices, including the Ace, Chao, Shannon, and Simpson indices, were calculated for all samples using Mothur (version 1.33.3) software. Weighted and unweighted uniFrac based on the Bray-Curtis distance were used to assess the beta diversity between samples. Meta stats were employed to determine diversity in intestinal Microbiome abundance between the MDD group and HCs.

Statistical analyses

Demographic and clinical data, the measurement data were expressed as mean ± SD, and the count data were expressed as frequencies. The t-test is used for data obeying a normal distribution, while the Mann-Whitney U test is applied to non-normally distributed data. The Wilcoxon test was used to study the relationship between fecal microbiota composition and MDD. Furthermore, false discovery rate (FDR) correction was carried out for multiple comparisons. Based on nonparametric tests, OTU cluster analysis was performed to analyze the differences between the MDD and HC groups at the OTU level. Spearman’s correlation analysis was used to explore the relationship between HAMD scores and various taxonomic units. Applying partial correlation analysis to remove the confounding effects of both age and gender variables to test the association of Hcy and cognitive functioning. We used the R studio software to analyze the data and graph with ggpubr, rstatix, patchwork, ggplotify, and vegan packages. Additionally, the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 (PICRUST2) is used to analyze the potential functional pathways for the two groups. The metagenome Kyoto encyclopedia of genes and genomes (KEGG) orthologs we detected were mapped to pathway level (KEGG level 2), with Wilcox. test to compare the MDD and HC groups. The results of all experiments were considered statistically significant at P < 0.05.

RESULTS

Demographic characteristics

Sixty-seven patients with MDD and ninety-four HCs participated in our research. There were no obvious differences in age, gender, years of education, and BMI between the two groups (P > 0.05; Table 1). Serum Hcy levels were significantly increased in the MDD group than in HCs (P < 0.001). The total HAMD score in depression was also increased than in HCs (P < 0.001). Among the seven dimensions of the MCCB, the total scores in patients with MDD were much decreased than that of the HCs, especially in terms of SOP, WM, BVMT, MAZES, and social EIT (P < 0.001). HVLT scores differed between the two groups (P = 0.01). Only CPT scores did not show clear differences between the two groups.

Table 1 Demographic characteristics and clinical data of patients with major depressive disorder and healthy controls, mean ± SD.

Characteristic	MDD (n = 67)	HCs (n = 94)	t or χ2	P value
Age (years)	34.97 ± 14.01	35.46 ± 11.44	0.23	0.82
Sex (male/female)	30/37	32/64	1.48	0.22
Education (years)	13.97 ± 2.82	14.80 ± 3.73	1.59	0.12
Body mass index (kg/m2)	23.97 ± 3.89	23.14 ± 3.57	-1.39	0.17
Duration (years)	5.09 ± 6.25	NA	NA	NA
Hcy (μmol/L)	20.94 ± 15.88	14.28 ± 4.17	-3.35	0.001b
HAMD-17 scores	22.37 ± 6.19	0.66 ± 1.41	-28.18	< 0.001c
MATRICS consensus cognitive battery
Speed of processing	53.57 ± 9.94	63.63 ± 13.16	5.52	< 0.001c
Attention/vigilance	50.99 ± 8.78	53.32 ± 8.10	1.74	0.08
Working memory	54.49 ± 8.49	61.27 ± 9.81	4.56	< 0.001c
Verbal learning	53.09 ± 9.44	57.24 ± 11.01	2.5	0.01a
Visual learning	48.40 ± 10.51	55.13 ± 9.29	4.28	< 0.001c
Reasoning and problem-solving	51.04 ± 10.15	58.39 ± 5.69	5.36	< 0.001c
Social cognition	45.16 ± 11.31	53.77 ± 10.66	4.92	< 0.001c
Total score	51.33 ± 8.56	59.73 ± 7.80	6.47	< 0.001c

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

Hcy: Homocysteine; HAMD: Hamilton depression rating scale; MDD: Major depressive disorder; HC: Healthy control; NA: Not applicable.

Intestinal flora

The α-diversity between the groups was significantly different (P < 0.05). Specifically, the Chao and Ace indices of the MDD group were significantly lower than HCs. Involving Shannon and Simpson indices, which are not clearly distinguishable between the two groups (Figure 1). Principal coordinate analysis of the Bray-Curtis distance index indicated that, based on the β-diversity, the structure of the intestinal microbial community was significantly different between the MDD and HCs groups (Figure 2). To investigate the differences in abundance between the two groups and further explore the specific bacteria associated with MDD. We determined the differential microbiota between the MDD and HCs groups by taxonomic comparison. The results revealed the top 20 bacterial groups in the two groups (Figure 3A), of which 11 OTUs were primarily distributed at the genus level. There were more relative abundances of Bacteroides, Escherichia, Parabacteroides, and Lachnospiraceae in the MDD than in the HCs group (P < 0.01). Compared with HCs, the relative abundances of the other nine genera (Prevotella, Faecalibacterium, Bacteroides, Megamonas, Lachnospiraceae, Lachnospira, Subdoligranulum, Dialister, Blautia) decreased (P < 0.01).

Figure 1 The alpha diversity indices between healthy controls and major depressive disorder groups. A: Chao; B: Ace; C: Shannon diversity; D: Simpson diversity. Wilcoxon signed-rank test was applied to compare the alpha diversity indices between healthy controls and major depressive disorder groups. MDD: Major depressive disorder; HC: Healthy control.

Figure 2 The beta diversity indices between healthy controls and major depressive disorder groups. The beta diversity distance matrix is displayed using principal coordinate analysis graphs to compare the sample distribution between healthy controls and major depressive disorder groups. Green squares and red points represent major depressive disorder patients and healthy controls, respectively. PCoA: Principal co-ordinates analysis; PC1: Principal component 1; PC2: Principal component 2; MDD: Major depressive disorder; HC: Healthy control.

Figure 3 Relative abundance. A: Comparison of relative abundances between healthy controls and major depressive disorder groups. The top 20 operational taxonomic units with the highest abundance and differences were selected; B: The top 20 functional pathways with the highest abundance and differences. The functional pathways were selected using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. ^dP < 0.0001. OTU: Operational taxonomic units; MDD: Major depressive disorder; HC: Healthy control.

Function prediction with PICRUST2

We performed functional predictive analyses based on fecal microbiological assays using the KEGG database.

In pathway level 2, we observed 29 significantly different pathways after FDR correction (P < 0.05). We selected the top 20 pathways in the two groups (Figure 3B). The identified metabolic pathways contained carbohydrate metabolism, amino acid metabolism, translation, replication and repair, membrane transport, glycan biosynthesis and metabolism, etc.

Relationship between the intestinal flora and clinical symptoms

The total HAMD score in the MDD group was negatively correlated with the gut microbiota (P < 0.05), specifically, Ruminococcaceae [r = -0.4, P < 0.001, 95% confidence interval (CI): -0.590 to -0.167], Faecalibacterium (r = -0.26, P = 0.036, 95%CI: -0.455 to -0.024), and Holdemanella (r = -0.26, P = 0.037, 95%CI: -0.445 to -0.062) (Figure 4).

Figure 4 Correlation between Hamilton depression rating scale and gut microbiota. A: Relationships between Hamilton depression rating scale (HAMD) scores and the relative abundance of Ruminococcaceae; B: Relationships between HAMD scores and the relative abundance of Faecalibacterium; C: Relationships between HAMD scores and the relative abundance of Holdemanella. The correlations were evaluated using Spearman correlation analysis. OTU: Operational taxonomic units; HAMD: Hamilton depression rating scale.

Associations between Hcy and cognitive function

To investigate the correlation between Hcy and cognitive functioning, confounding (age and gender) was controlled for using partial correlation analyses, and this experiment primarily controlled for the effects of age and sex (Figure 5). In the MDD group, Hcy showed a negative correlation with total MCCB score (r = -0.249, P = 0.046, 95%CI: -0.465 to -0.05) and SOP index (r = -0.367, P = 0.003, 95%CI: -0.561 to -0.135). Hcy did not show significant differences across HVLT, CPT, WM, MAZES, EIT, and BVMT scores (P > 0.05). Among the HCs, Hcy levels were only negatively correlated with the MCCB and EIT index scores (r = -0.244, P = 0.019, 95%CI: -0.428 to -0.041) and total scores (r = -0.222, P = 0.033, 95%CI: -0.408 to -0.018). There were no correlations with the other items.

Figure 5 Correlation between cognitive function and homocysteine in healthy controls and major depressive disorder groups. A: Partial correlation analysis between homocysteine (Hcy) and speed of processing in the major depressive disorder (MDD) group; B: Partial correlation analysis between Hcy and total score in the MDD group; C: Partial correlation analysis between Hcy and social cognition in healthy controls (HCs); D: Partial correlation analysis between Hcy and total score in HCs. Hcy: Homocysteine; SOP: Speed of processing; EIT: Social cognition; MCCB: MATRICS consensus cognitive battery.

Relationship between Hcy and gut microbiome

Exploring the association between Hcy and gut flora using Spearman’s correlation analysis. In both groups, Hcy levels were negatively correlated with cognitive function (Figure 6) (P < 0.05). In patients with MDD, Hcy was negatively correlated with Alistipes, Ruminococcaceae, Tenericutes, and Porphyromonas (P < 0.05) and positively correlated with Megasphaera (P < 0.05). In HCs, Hcy was negatively correlated with Coprococcus (P < 0.001) and negatively related to Pelomonas, Lachnospiraceae, and Tenericutes (P < 0.01). Simultaneously, it has a positive correlation with Lachnospiraceae (P < 0.01) and Escherichia (P < 0.05) (Figure 5).

Figure 6 Associations between serum homocysteine and intestinal flora. The associations in healthy controls and major depressive disorder groups. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. ^dP < 0.0001. OTU: Operational taxonomic units; MDD: Major depressive disorder; HC: Healthy control.

DISCUSSION

The results of our study firstly showed significant differences between the two populations in regard to Hcy levels, gut microbiota structure, cognitive function, and other clinical evaluation indicators. Elevated levels of Hcy, gut microbiome dysfunction, and poorer cognitive function were found in patients with MDD. Hcy was inversely associated with cognition, particularly processing speed. We also found that serum Hcy levels were inversely interrelated with the abundance of some intestinal flora, such as Alistipes, Ruminococcaceae, Tenericutes, and Porphyromonas in MDD. Additionally, our functional analysis indicated that carbohydrate metabolism, lipid metabolism, and amino acid metabolism showed significant differences between the MDD and HCs.

There is a correlation between Hcy and the development of multiple psychiatric disorders. A recent randomized Mendelian study demonstrated that elevated Hcy concentrations increase the risk of developing SCZ and bipolar I disorder[39]. A large sample study also found that Hcy was positively correlated with depressive symptoms[40]. Similarly, Folstein et al[41] published that elevated serum Hcy increases the risk of MDD and tends to indicate a poor prognosis. Consistent with previous findings, our findings confirm that patients with MDD tend to show higher levels of Hcy concentrations indeed. Excessively elevated may have neurotoxic effects and exacerbate oxidative stress to increase susceptibility to depression[40]. Folic acid supplementation can help reduce depressive symptoms[42,43]. This phenomenon may be related to the intricate interrelationship between folate, Hcy, and monoamine metabolism[44,45]. These observations collectively indicate that Hcy concentrations are relevant to the onset, development, and prognosis of depression, although the specific mechanisms remain unclear.

Cognitive dysfunction, one of the central symptoms of depression, is an essential mediator of dysfunction in depressive states[46]. Our study also found the cognition of patients with MDD declined, such as memory, learning ability, processing speed, EIT, and so on. A piece of evidence to support Hcy is potentially pathogenic for cognitive impairment[47]. Specifically, elevated Hcy levels are widespread in Alzheimer’s disease, a disease with severe cognitive impairment. A meta-analysis suggested a causal link between increased Hcy levels and increased risk of dementia[48]. All of the above studies showed a correlation between Hcy and cognition. The impact of Hcy levels on cognitive function is particularly pronounced in elderly patients with depression[21]. Another recent research showed that Hcy might play a potential role in the diagnosis and treatment of MDD[49]. In most review studies, patients’ memory and emotional states were affected by Hcy concentration. Our study found that Hcy was negatively correlated with the speed of processing in MDD, and the results confirmed Rutherford et al’s work[50]. We hypothesize that this association is related to B vitamins and folate metabolism, both are important coenzymes for 5-HT and NE synthesis. 5-HT2A and NE seem to be sufficient to explain the altered state of mood and cognitive function[51,52].

The latest study demonstrated that an increase in Hcy alters the consisting of intestinal bacteria and reduces the richness of Lachnospiraceae and Alistipes[26]. It was also confirmed in our experiment. When serum Hcy was maintained at a high level, there existed obvious diversity in gut microbes composition between MDD and HCs. Specifically, the fecal samples of the former had higher levels of Bacteroides, Proteobacteria, and Actinomycetes[53]. Our results suggest that the abundance of Alistipes and Porphyromonas were negatively correlated with Hcy, both of which belong to the Bacteroidetes. It was published that excessive serum Hcy disrupts the intestinal barrier, leading to intestinal microbial dysbiosis[54,55]; however, Hcy levels decrease significantly after the recovery of microbiota homeostasis[27], suggesting a strong association between Hcy and gut microbiota. In a recent study on SCZ, serum Hcy levels had a negative correlation with cognition and were positively correlated with intestinal beneficial bacteria, including Eubacterium and Lactobacillus. These probiotics are closely related to amino acids and metabolism[31]. In addition, gut microbiome depletion could influence the metabolism of methionine, further causing a decrease in Hcy and S-adenosine methionine[56]. All above explained the potential link between Hcy and microbiome, and our results also innovatively found the relationship in MDD, especially between Hcy and Bacteroidetes. However, the direct relationship between the two in patients with MDD remains inconclusive.

There is a possible association between Hcy, mental disorders, and cognitive function, while current studies tend to suggest that the relationship may be related to the role of gut microbiota. It is hypothesized that alterations in the composition of the gut microbiota may play a pivotal role in the pathogenesis of cognitive impairment in individuals with MDD. Gut flora can produce a variety of metabolites by metabolizing proteins and amino acids. Short-chain fatty acids (SCFAs), such as butyric, propionic, and acetic acids, are the most prevalent metabolites. These metabolites can influence the availability and metabolism of amino acids in the individual. They also affect the synthesis and release of monoamine neurotransmitters[57,58], which may affect mood and cognitive function[55,59]. The Bacteroides and Firmicutes are often considered to be strongly associated with cognitive impairment. Several studies have demonstrated that both phyla frequently exhibit aberrant proportions and composition in individuals with cognitive impairment[23,60,61]. Bacteroidetes belong to the group of anaerobic, gram-stain-negative bacteria, which are the major producers of butyrate in intestinal bacteria[62]. Previous studies have shown that Bacteroidetes are associated with neurodevelopment and positively correlate with cognitive and language scores during infancy[63]. Firmicutes are also known to produce SCFAs in significant quantities. However, it has been postulated that their excessive increase may be associated with the onset and development of cognitive impairments[64,65]. Our findings also indicate a negative correlation between gut microbiota and EIT in healthy individuals, primarily due to the role of the phylum Firmicutes (Coprococcus, Lachnospiraceae). Dysbiosis may be present in patients with MDD and normal individuals. This dysregulation may affect cognition in the two groups in different ways. In normal individuals, the dysregulation may affect multiple aspects of cognitive function, while in patients with depression, its effects may be relatively minor. Furthermore, it has been demonstrated that acetate interventions influence cognition in mice[66]. The available evidence suggests a correlation between changes in the gut microbiome and cognitive deficits in depression, with the metabolites of Bacteroidetes being the most likely underlying cause. It may be possible to follow up by constructing animal models to determine Hcy-driven changes in SCFAs in the gut and their association with cognitive symptoms of depression.

Emerging evidence increasingly supports the role of dietary and lifestyle factors in regulating gut microbiota and Hcy levels. Healthy lifestyle interventions, including regular physical exercise, mindfulness practices, meditation, and stress-relief techniques such as music therapy, have been shown to help maintain the stability and diversity of the gut microbiota[67-69]. Previous studies have reported that moderate-intensity exercise is considered the most beneficial for gut microbiota diversity, which may be related to the increased abundance of SCFA-producing bacteria[70,71]. However, the specific mechanisms underlying this relationship remain unexplored. The diet also plays a crucial role in gut health[72]. Specific dietary patterns influence gut barrier function and systemic inflammatory responses by modulating the composition of the gut microbiota and its metabolites, ultimately affecting host health. The Mediterranean and anti-inflammatory diets are two popular dietary patterns that are believed to improve gut microbiota structure and promote the growth of beneficial bacteria, including Bifidobacterium and Lactobacillus. The Mediterranean diet is characterized by its high fiber content and abundance of polyphenols[73], which enhance the production of SCFAs, reduce gut permeability, and strengthen intestinal health[74,75]. Similarly, the anti-inflammatory diet emphasizes the intake of omega-3 fatty acids, antioxidants, and dietary fiber, which can further regulate microbial metabolites and improve systemic immune responses[76]. Tryptophan in the diet is converted into kynurenine and indole compounds in the gut. Gut bacteria and enzymes such as Indoleamine 2,3-dioxygenase drive this conversion. Certain probiotics, including Prevotella and Bacteroides, can boost indole production from tryptophan, which helps modulate the kynurenine pathway, reducing inflammation and potentially improving mood. In contrast, a high-fat diet negatively impacts the gut microbiota by promoting the growth of pro-inflammatory bacteria, increasing gut permeability, and inducing systemic inflammation[77]. These inflammatory processes are closely associated with elevated Hcy levels.

Additionally, specific dietary nutrients can influence and regulate the gut-brain axis through metabolic pathways. A methionine-rich diet is considered a significant risk factor for elevated Hcy levels. In the body, methionine is metabolized into Hcy. Elevated Hcy levels can increase SAM levels, which subsequently inhibit the remethylation of Hcy, causing further accumulation of Hcy. This metabolic imbalance triggers a series of oxidative stress reactions, negatively affecting the normal synthesis and metabolism of neurotransmitters. Vitamin B6, B12, and folate are important coenzymes for methionine, and folate cycles can impact Hcy levels[78]. Studies have shown that exogenous supplementation with B vitamins and folate can accelerate Hcy’s metabolism and effectively reduce its plasma levels. A previous large-scale review indicated that, in addition to nutrient intake, maintaining healthy dietary habits (such as increasing the consumption of fruits and vegetables) and adopting a healthy lifestyle (such as regular exercise) can help maintain lower Hcy levels. These interventions contribute to improving metabolic balance and enhancing the body’s antioxidant capacity, thereby supporting the maintenance of low Hcy levels[79]. However, in our study, all participants were from Beijing and shared similar dietary habits. The patients surveyed were all inpatients, and the hospital cafeteria provided their meals. Their lifestyle, daily routines, and levels of physical activity were also highly consistent. Therefore, we did not conduct a detailed investigation of the influence of individual dietary and lifestyle factors.

Our findings highlight the intricate relationships among gut microbiota, Hcy, and cognitive function, unveiling potential biological connections among these factors and offering new perspectives on the mechanisms underlying the gut-brain axis. We have confirmed that there is a certain correlation between the microbial community and cognition, and studies have also shown that supplementing with probiotics can simultaneously improve cognitive function and depressive mood[80]. A previous large-sample follow-up study suggested that the SAM cycle I pathway offers a plausible mechanism by which the microbial community is associated with depressive disorders[14]. We also speculate that the gut microbiota may be the key medium linking Hcy with cognition, and the underlying mechanism is related to the methylation cycle. Our findings lay the foundation for further research. However, further validation is still needed. By regulating the gut microbiome, lowering Hcy levels, reducing inflammation, and providing antioxidant protection, there is potential for adjunctive therapeutic strategies for depression. This suggests that targeting gut microbiota could serve as a novel approach to improving depressive symptoms and cognitive function. Furthermore, these findings provide a scientific basis for developing microbiota-based interventions, including probiotics or dietary modifications, offering a more comprehensive strategy for managing depression.

Our study is highly novel, featuring a comprehensive assessment of gut microbiota composition, cognitive performance, and Hcy levels, as well as a multidimensional analysis of their interconnections. However, our study had some limitations indeed. First, as a cross-sectional study, it does not allow for conclusions about causal relationships, leaving the precise interaction mechanisms in patients with MDD unclear. Second, compared with metagenomics, the accuracy of 16s rRNA detection is still low because 16s is mainly studied in bacteria, and the metagenome can be better analyzed at the gene level. Considering the effect of microbiota in the onset and progression of diseases in all systems, fecal metabolism correlation analysis and other means can be added to study the specific mechanisms further. Then, we did not conduct corresponding validation experiments, such as cellular or in vivo animal studies; thus, our findings remain exploratory hypotheses rather than conclusive mechanisms. Additionally, our participants were inpatients from the same region with similar dietary habits and maintained comparable lifestyles, sleep schedules, and physical activity levels due to hospitalization. However, we did not collect detailed data on diet and lifestyle, which represents a limitation that we plan to address by including relevant questionnaires in future research. Finally, all participants were from the North China region, which may limit the generalizability of our findings. More importantly, this regional homogeneity might fail to fully capture the diversity of microbial strains. To address these limitations, multicenter studies involving larger and more diverse populations should be conducted to validate and extend our findings. Future studies should incorporate long-term follow-ups and, where appropriate, combine relevant foundational research and mediation effect analyses to further explore its underlying mechanism.

CONCLUSION

We experimentally demonstrated a potential link between serum Hcy, gut microbiota, and cognitive impairment in patients with MDD. It was also characterized by characteristic changes in their serum Hcy levels and gut microbiota. The findings suggest that elevated serum Hcy levels are negatively correlated with cognitive function in patients with MDD and that this relationship may be related to changes in the structure of the intestinal microbial community. However, factors such as individual variability, gut flora interactions, and the overall health status of patients may influence the experimental results. Therefore, the exact mechanism needs to be further investigated. In conclusion, our findings align with previous studies and suggest that the gut-brain axis may play a significant role as a potential pathological mechanism of depression. Supplementing specific strains of bacteria may, therefore, be an effective method of improving cognitive function in patients, which could have a positive impact on the prognosis of the disease and restoration of social functioning. This provides an effective basis for advancing the treatment of gut-mediated depression.

ACKNOWLEDGEMENTS

We gratefully acknowledge all patients and researchers who participated in this study.