Does isovaleric acid play a key role in the interaction between probiotics and antidepressants? A secondary analysis of a randomized clinical trial

Abstract

Background

Dysbiosis appears to be a significant contributor to the complex pathophysiology of mood disorders, and short-chain fatty acids (SCFAs), the main metabolites produced in the colon by bacterial fermentation, have been found to play a role in gut-brain communication. Probiotics were shown to be effective in managing and alleviating depressive symptoms, especially as an add-on protocol. This study aimed to assess the change in fecal SCFAs levels after supplementation with probiotics in patients with depression, depending on baseline antidepressant treatment.

Methods

This was a secondary analysis of a two-arm, parallel-group, randomized, double-blind, controlled trial. Data from 65 participants were analyzed. The intervention included probiotic formulation (Lactobacillus helveticus Rosell®-52 and Bifidobacterium longum Rosell®-175; R0052/R0175) or placebo over a 60-day period. Then, stratification was performed by the type of antidepressant medications. Fecal SCFAs were measured by the gas chromatography method. Pre-intervention socio-demographic, clinical, and laboratory data were assessed.

Results

Probiotics used decreased the levels of isovaleric acid compared with placebo when administered with non-selective serotonin reuptake inhibitors antidepressants (non-SSRIs) with large effect size (p = .019, |r|=.653), but not when used with SSRIs (p = .572, |r|=.109) or applied alone (p = .404, |r|=.182). Isovalerate levels decreased as depression improved in the probiotic plus non-SSRIs group. Conclusions: R0052/R0175 as an add-on to non-SSRI antidepressants may offer antidepressant action partly through the decrease in isovaleric acid levels. More research with a larger sample size is needed to study SCFAs' role as a mediator of antidepressant action of both probiotics and medications. ClinicalTrials.gov identifier: NCT04756544.

Graphical Abstract

Highlights

• •Probiotics did not change fecal short-chain fatty acid levels overall.
• •C5i levels were decreased after probiotics plus non-SSRI antidepressants.
• •C5i levels decreased as depression improved in the PRO plus non-SSRIs group.
• •Increased C2 levels were associated with depression improvement overall.
• •Decreased C3 levels were associated with depression improvement overall.

1. Introduction

Depression is a widespread and debilitating neuropsychiatric disorder that impairs behavior, feelings, thoughts, and overall well-being. According to a 2024 report by the World Health Organization (WHO), approximately 5 % of adults worldwide suffer from depression, making it the third leading cause of disability globally. However, this prevalence can vary significantly between countries and regions, often being higher in more developed nations. In particular, lifetime diagnoses of depression paint a more extensive picture. For example, in the United States, 18.4 % of adults report having been diagnosed with depression at some point in their lives [1]. The prevalence of depression also varies across different age groups. A recent review of 48 studies estimated the global prevalence of depression in older adults to be 28.4 % [2]. Meta-analyses indicate that the prevalence of all depressive disorders, particularly during the COVID-19 pandemic, reached even higher levels, ranging from 37 % to as much as 45 % in some studies. This includes both major and minor depressive conditions [3].

The etiology of depression is multifactorial, involving biological, genetic, environmental, and psychosocial factors. The condition was initially attributed to abnormalities in neurotransmitters such as serotonin (5-HT), norepinephrine (NE), and dopamine (DA); this view is supported by the effectiveness of antidepressants targeting these pathways. However, recent theories suggest depression arises from complex neuroregulatory and neural circuit disturbances, with secondary effects on neurotransmitter systems [4]. However, the role of gut microbiota in the pathophysiology of depression has garnered growing interest. Dysbiosis, the imbalance in the bacterial microbiota, appears to be a significant contributor to the complex pathophysiology of mood disorders [5], [6]. Many microbial metabolites can cross the blood–brain barrier and have significant effects on the brain, playing a key role in the so-called microbiota-gut-brain axis (MGBA). Dysregulation of the MGBA is now recognized as a key feature of various neuropsychiatric conditions, including depressive disorder [7], [8]. Notably, alterations in MGBA function are often linked to changes in bacterial metabolites, such as short-chain fatty acids (SCFAs), which have profound effects on brain function [9]. Furthermore, gut microbiota is not only essential for maintaining overall health but is also closely linked to the efficacy and side effects of drugs, including antidepressants.

SCFAs are end-products of intestinal microbial fermentation of non-digestible carbohydrates in the gut. The main SCFAs produced by the gut microbiota are acetic acid (C2), propionic acid (C3), butyric acid (C4n), valeric acid (C5n), and caproic acid (C6n). They play a crucial role in gut-brain communication and maintaining colonic health. With butyrate as a key player, SCFAs act as a preferred fuel for colonocytes, influencing cell proliferation and apoptosis, and regulating tight junction proteins to enhance intestinal barrier function. Additionally, SCFAs play a multifaceted role in regulating glucose homeostasis, impacting hepatic processes and adipose tissue [10], [11]. Furthermore, SCFAs exhibit immunomodulatory potential by various mechanisms, including stimulation of G protein–coupled receptors, inhibition of histone deacetylases (and hence, gene transcription changes), increases in regulatory T-cell numbers and function, and decreased expression of numerous inflammatory cytokines, offering therapeutic promise in inflammatory conditions [12]. One of the largest meta-analyses to date exploring the association between SCFAs and mood disorders, including 11 human studies with 662 participants and 330 controls, found that while most studies reported associations between SCFA levels and depressive symptoms, the directionality and specific SCFAs varied [13]. Notably, recent studies have highlighted different concentrations of SCFAs across different biological matrices: fecal and plasma levels often show opposing changes in patients compared to healthy controls [14]. This inversion can arise from altered gut epithelial transport and barrier function, whereby increased colonic uptake depletes fecal SCFAs even as systemic concentrations rise [15].

Overall, shifts in the ratio of SCFAs are more commonly associated with depressive symptoms than their absolute levels. Supporting this, a previous review including 26 studies on anxiety disorders and depression also found that patients had a lower abundance of short-chain fatty acid-producing species compared to controls [16]. Simpson et. al. further noted that although findings on alpha and beta diversity were inconsistent, bacterial taxa associated with these disorders consistently revealed a higher abundance of proinflammatory species (e.g., Enterobacteriaceae and Desulfovibrio) and a lower abundance of SCFA-producing bacteria (e.g., Faecalibacterium). Moreover, among depressed patients, negative correlations were observed between stool SCFAs concentrations and scores on the Beck's Depression Inventory (BDI). These findings underscore the potential role of gut microbiota composition and SCFA metabolism in the pathophysiology of mood disorders [17].

In addition to the aforementioned acids, the microbiota also produces considerably lower amounts of a subclass of saturated fatty acids commonly known as branched short-chain fatty acids (BCFA), including isobutyric acid (C4i), isovaleric acid (C5i), and isocaproic acid (C6i). In the human intestine, the fermentation of BCFAs is carried out mainly by the genera Bacteroides and Clostridium [18]. BCFAs levels increase from the proximal to the distal colon, with fecal BCFAs levels serving as markers of colonic protein fermentation. Dietary changes, such as complex carbohydrate supplementation, reduce BCFAs levels, while protein supplementation leads to increased production [19]. While SCFAs like butyrate and propionate have been extensively studied, the roles of BCFAs remain poorly understood, and the results are mixed, despite growing evidence suggesting their involvement in gut health, systemic inflammation, and mental health [20].

First of all, Szczesniak et al. analysed stool samples from 34 patients with major depression and 17 healthy controls, measuring a range of volatile fatty acids (VFAs) alongside salivary cortisol [21]. The researchers observed a pronounced bimodal distribution of isovaleric acid, with depressed patients significantly overrepresented in the high–isovaleric subgroup. Moreover, stool isovaleric acid levels positively correlated with salivary cortisol concentrations, suggesting that increased gut-derived C5i may be linked to hypothalamic-pituitary axis (HPA) activation in depression.

As regards plasma BCFAs, Yu et al. conducted a cross-sectional study on patients with major depressive disorder (MDD), schizophrenia, and healthy controls, in which they measured the levels of inflammatory mediators and VFAs [22]. Concentrations of caproic, isovaleric, isobutyric, propionic, acetic, and valeric acids were negatively correlated with depression, anxiety, and both negative and general psychopathology scores of schizophrenia. Integrating the top discriminative SCFAs and inflammatory markers into a predictive model accurately distinguished MDD from controls (AUC = 0.997). The model included C6, C5i, C5, and C3, in order of their importance. In another study, Yu et al. found that total plasma VFAs — and specifically acetic, propionic, isovaleric, isobutyric, and valeric acids—were significantly reduced in patients with bipolar depression before and after treatment with quetiapine compared to controls [23]. Correlation analyses revealed that lower levels of C3, C5, and C5i were associated with greater depressive symptom severity, highlighting their potential as state-related biomarkers in bipolar depression.

Furthermore, Pu et al. (2022) conducted a vote-counting meta-analysis of 157 animal studies to identify metabolites consistently altered by pharmacological treatments in depression models, including a focus on gut microbiota–derived substances. They found that antidepressant interventions not only normalized central neurotransmitter and amino acid disturbances but also significantly increased fecal isovaleric acid, alongside acetate, propionate, and butyrate [24].

To sum up, high fecal and low circulating levels of isovaleric acid (C5i) may be associated with human depression or its treatment. These findings underscore the potential relevance of C5i as one of the biomarkers of dysbiosis linked to depression occurrence and severity. Nonetheless, the physiological significance of isovalerate is not fully known.

For simplicity, as the article progresses, both major short-chain fatty acids (acetate, propionate, and butyrate) and branched-chain fatty acids will be referred to collectively as SCFAs, as BCFAs are often considered a subclass of SCFAs.

Accumulating studies have demonstrated the beneficial effects of gut microbiota management tools in improving depression. Among these, probiotics, specifically selected microorganisms that provide health benefits to the host, have shown significant effectiveness in managing and alleviating depressive symptoms. A recent meta-analysis of 13 randomized controlled trials with 786 participants showed that prebiotics, probiotics, or synbiotics significantly improved depressive symptoms compared to placebo [25]. Emerging evidence also indicates that the link exists between gut microbiota interventions and changes in SCFAs, with a recent meta-analysis finding that pre- and probiotic treatments increased butyrate levels and improved depression scores [26]. Thus, probiotic interventions could serve as a safe and user-friendly supplementary treatment option for depression and its coexisting conditions. Therefore, it is crucial to conduct further well-designed clinical trials to better understand the specific mechanisms, particularly the role of SCFAs in mood regulation, as well as to establish standardized protocols for monitoring SCFAs levels and their correlation with clinical outcomes.

Importantly, probiotics appear to exert their antidepressant effects primarily as an add-on treatment alongside antidepressants [27]. The most often used antidepressant medications are selective serotonin-reuptake inhibitors (SSRIs), e.g. sertraline or fluoxetine. Non-SSRIs antidepressants include serotonin and norepinephrine-reuptake inhibitors (SNRI, e.g., venlafaxine), trazodone, clomipramine, and vortioxetine, among others. Some data exists on the effect of antidepressants on gut microbiota composition and some of the metabolites produced by microbiota (reviewed in [28]). For instance, a recent study found that patients with depression treated with SSRIs or SNRIs exhibited decreased microbiota alpha diversity and an altered microbiota composition compared with those with depression but not receiving antidepressant treatment [29]. Collectively, these findings suggest that dysbiosis could be a new therapeutic target and prognostic tool for the treatment of patients with depression. Nonetheless, very little is known about the influence of different antidepressant medications on SCFAs levels. Very little data exists regarding the impact of antidepressant treatment on the changes in C5i levels and no data has been acquired on the influence of the combination of probiotics and antidepressants on fecal SCFAs concentrations.

The study presents the results of a secondary analysis of the Pro-DEMET RCT [30], aimed at examining changes in fecal SCFA levels and ratios following probiotic intervention compared to placebo; the analysis was stratified according to the type of antidepressants used (Fig. 1).

Fig. 1.

The hypothesized interplay between microbiota-gut-brain-axis, fecal short-chain fatty acids (SCFAs), probiotics and antidepressants. Abbreviations: C2 – acetic acid; C2 – propionic acid; C4n – butyric acid; C4i – isobutyric acid; C5n – valeric acid; C5i – isovaleric acid; Il-6 – interleukin-6; TNF-α – tumor necrosis factor alpha.

The main objectives of our analysis were:

• (i)To compare the change in fSCFAs in the PRO (probiotic) with the PLC (placebo) group;
• (ii)To determine the change in fSCFAs (PRO vs. PLC) stratified by different antidepressant treatments (none / SSRIs / non-SSRI antidepressants);
• (iii)To correlate changes in psychometric scale scores with changes in fSCFAs stratified by different antidepressant treatments.

It is hypothesized that in patients with depression, supplementation of probiotics as an add-on therapy to antidepressants results in a change in fecal SCFAs levels. Moreover, the change in fecal fatty acids corresponds to the efficacy of probiotics regarding depressive symptoms.

2. Materials and methods

2.1. Design

The parent study was a parallel-group, randomized, double-blind, placebo-controlled trial performed at the Medical University of Lodz, Poland, between December 2020 and August 2023 [31]. Adult patients with depressive disorders according to the 11th International Classification of Diseases [32] were randomly assigned to groups via computer-generated blocked lists stratified by the presence of MetS. They underwent 60 days of supplementation with probiotics (Lactobacillus helveticus Rosell®-52 and Bifidobacterium longum Rosell®-175) or placebo daily. Intervention details are shown in the Supplement. Randomization was performed using a computer-based random number generator (https://www.randomizer.org/, accessed on December 10th, 2020).

The presented analysis sample finally consisted of 65 subjects. Inclusion and exclusion criteria, and the study timeline, were presented previously [30], [31] and are described in the Supplement.

This study manuscript has been planned and prepared according to the CONSORT statement for RCTs [33].

Table 1 shows the outcome measures. The change in isovaleric acid (C5i) was selected as this study's primary outcome measure based on its emerging significance as a marker of colonic protein fermentation and its potential role in gut-brain communication.

Table 1.

The analysis outcome measures.

Outcome measures
Primary	∆ isovaleric acid (C5i)
Secondary	Δ acetic acid (C2), Δ propionic acid (C3), Δ butyric acidC(4 n), Δ isobutyric acid (C4i), Δ valeric acid (C5n)
Tertiary	Baseline: antidepressant treatment, I-FABP, bSCFAS, dietary habits, physical activity level, MADRS, DASS, QoL, MCID MADRS %Δ MADRS, %Δ DASS, %Δ D-DASS, %Δ A-DASS, %Δ S-DASS

Legend: Δ – change between the end (V2) and the beginning (V1) of the intervention period; %Δ – percentage Δ. Abbreviations: A-DASS – Anxiety subscale of DASS; bSCFAs – blood short-chain fatty acids; DASS – Depression, Anxiety, and Stress Scale; D-DASS – depression subscale of DASS; I-FABP – intestinal fatty acid-binding protein; MADRS – Montgomery-Asberg Depression Rating Scale; MCID – minimal clinically important difference; S-DASS – Stress subscale of DASS; QoL – Quality of Life.

2.2. Methods

2.2.1. Questionnaires

The study-specific questionnaires were used to capture demographic, lifestyle, or health-related data, and monitor them throughout the intervention period [30]. Dietary habits were assessed with the Food Frequency Questionnaire-6 [34], physical activity with the International Physical Activity Questionnaire [35], negative emotional states with the Montgomery-Asberg Depression Rating Scale (MADRS) [36], or Depression, Anxiety, and Stress Scale [37], quality of life with the WHOQoLBREF Instrument [38]. The characteristics of the questionnaires are published elsewhere [39].

2.2.2. Biological samples

Fasting venous blood was collected, and basic blood test were performed in the Department of Laboratory Diagnostics, Central Teaching Hospital, Medical University of Lodz, Poland. The blood serum was stored at –80°C for further analyses. Intestinal fatty acid-binding protein (I-FABP) and blood short-chain fatty acids (bSCFAs) were measured using commercial ELISA kits at the Department of Biomedicine and Genetics, Medical University of Lodz, Poland. The measurements of propionic acid (PA) and citrulline (CIT) levels were performed in the Institute of Food Technology and Analysis, Lodz University of Technology, Poland using a Shimadzu GC-MS (Shimadzu, Tokyo, Japan) or by UHPLC-ESI-MS using the MetAmino® Kit (Chromservis, Czech Republic).

Stool samples were self-collected into sterile containers after overnight fasting twice, at the beginning (V1) and the end (V2) of the intervention period. No use of laxatives, synthetic fat substitutes, or fat-blocking supplements was allowed within 24 hours before sampling. The samples were frozen on the same day and stored at −80◦C in the BioBank of the Medical University of Lodz until further analyses. The following SCFAs were evaluated: acetic acid (C2), propionic acid (C3), butyric acid (C4n), valeric acid (C5n), caproic acid (C6n), and heptanoic acid (C7), as well as: isobutyric acid (C4i), isovaleric acid (C5i), isocaproic acid (C6i). The measurements were performed at the Department of Human Nutrition and Metabolomics, Pomeranian Medical University, Szczecin. The details of fecal SCFAs analyses are described in the Supplement.

2.2.3. Intervention

The probiotic (PRO) group received Lactobacillus helveticus Rosell®-52 and Bifidobacterium longum Rosell®-175 (3 ×10⁹ CFU), and excipients (Sanprobi Stress®, Sanprobi Sp. z o. o., Sp. k., Szczecin, Poland; Institute Rosell-Lallemand, Montreal, Canada). The placebo (PLC) group received only the excipients (Sanprobi Sp. z o. o., Sp. k., Szczecin, Poland).In both cases, the intervention proceeded for 60 days. For clarity, the term R0052/R0175 will be used for the formulation throughout the manuscript.

2.2.4. Data analysis

Statistical analyses of data were conducted in R (version 4.2.2). Results were considered statistically significant at p < 0.05.

For baseline characteristics, the Shapiro–Wilk test was used to check the distribution of data. If the data had normal distribution, Levene’s test was used to check the equality of variances, and Student’s t-test to determine statistically significant differences between pairs of groups. The data was presented as mean ± SD. If data did not have a normal distribution, the Wilcoxon-Mann-Whitney test was used to determine significant differences between pairs of groups, and the data was presented as median (IQR). Categorical data was analysed using Pearson's chi-squared test with the datapresented as a number (%).

The difference in the level (∆) of particular fSCFAs was calculated as (V2-V1). Asmost variables were non-normally distributed, the Wilcoxon-Mann-Whitney test was used to compare pairs of groups, and the Kruskal-Wallis test with post hoc analysis was used to compare multiple groups. Data was presented as median (IQR). Correlation analysis was performed using Spearman’s rank correlation test.

As multiple outcome measures emerged, one primary outcome (the change in C5i% levels) was chosen. Additionally, point estimates with 95 % confidece intervals (95 % CI) and effect size measures where used [40].

3. Results

3.1. Population characteristics

Fecal samples from both V1- and V2-time-points for 61 participants were available according to Per-Protocol analysis. Table 2 presents baseline characteristics of the study sample.

Table 2.

Characteristics of the study participants at the start of the trial. Data is shown as ratio, number (percentage), mean ± standard deviation, or median (interquartile range). N = 61, if not otherwise stated.

Characteristics	Total (N = 61)	PRO group (n = 28)	PLC group (n = 33)	p
Sex (F:M)	53:8	24:4	29:4	1
Age (y)	33.8 (22.5–41.9)#	32.3 (22.4–42.4)#	33.8 (24.1–40.1)#	0.948
Diagnosis according to ICD−11 (6A70/6A71:6A73)	21:40	14:14	7:26	0.178
Psychotropic medications (%)	40 (65.6 %)	16 (57.1 %)	24 (72.7 %)	0.314
Antidepressants (%)	40 (65.6 %)	16 (57.1 %)	24 (72.7 %)	0.314
SSRIs	27 (44.3 %)	11 (39.3 %)	16 (48.5 %)	0.644
Non-SSRIs	13 (21.3 %)	5 (17.9 %)	8 (24.2 %)	0.769
Fluoxetine dose equivalents (mg) (n = 32)	22.2 (20.3–43.1)	31.2 (20.3–47.2)	22.2 (20.3–40.2)	0.638
Antipsychotics (%)	4 (6.6 %)	3 (10.7 %)	1 (3 %)	0.491
Mood stabilizers	6 (9.8 %)	3 (10.7 %)	3 (9.1 %)	1
Other psychotropicmedications	5 (8.2 %)	3 (10.7 %)	2 (6.1 %)	0.848
Psychotropic polypharmacotherapy (%)	18 (29.5 %)	9 (32.1 %)	9 (27.3 %)	0.229
Comorbidities (%)	31 (50.8 %)	16 (57.1 %)	15 (45.4 %)	0.514
Other than psychotropic pharmacological treatment (%)	31 (34.3 %)	12 (42.9 %)	9 (27.3 %)	0.314
Smoking cigarettes (%)	9 (14.8 %)	4 (14.3 %)	5 (15.2 %)	1
Dietary supplements (%)	32 (52.5 %)	18 (64.3 %)	14 (42.4 %)	0.148
Obesity (according to BMI) (%)	5 (8.2 %)	1 (3.6 %)	4 (12.1 %)	0.456
Abdominal obesity (%)	35 (57.4 %)	17 (60.7 %)	18 (54.5 %)	0.821
Metabolic syndrome (%)	13 (21.3 %)	7 (25 %)	6 (18.2 %)	0.738
Chronic low-grade inflammation (%)	15 (24.6 %)	7 (25 %)	8 (24.2 %)	1
Physical activity (MET-min/week) (n = 17)	2108 ± 1615$	2755 ± 1960$	1656 ± 1235$	0.175
Fecal SCFAs (%)
Acetic acid (C2)	67.9 ± 8.54$	68.7 ± 10.4$	67.3 ± 6.66$	0.507
Propionic acid (C3)	0.68 (0.24–1.9)#	0.47 (0–1.91)#	0.77 (0.33–1.49)#	0.472
Butyric acid (C4n)	13.1 ± 7.12$	13.4 ± 8.26$	12.9 ± 6.12$	0.809
Isobutyric acid (C4i)	15.8 (12.9–19.8)#	14.4 (12.5–17.9)#	16.9 (13.6–21.2)#	0.089
Valeric acid (C5n)	0 (0−0)#	0 (0–0.02)#	0 (0−0)#	0.672
Isovaleric acid (C5i)	0.05 (0–1.72)#	0 (0–1.91)#	0.14 (0–1.47)#	0.793
Fecal SCFAs (mM)
Total	4.24 (3.37–6.98)#	4.03 (3.38–6.79)#	4.54 (3.35–6.98)#	0.914
Acetic acid (C2)	3 (2.23–4.3)#	2.9 (2.21–4.73)#	3 (2.27–4.25)#	0.937
Propionic acid (C3)	0.03 (0.01–0.08)#	0.02 (0–0.08)#	0.04 (0.02–0.08)#	0.365
Butyric acid (C4n)	0.57 (0.28–0.97)#	0.64 (0.27–1.21)#	0.41 (0.31–0.83)#	0.757
Isobutyric acid (C4i)	0.72 (0.5–1.19)#	0.64 (0.49–1.21)#	0.82 (0.52–1.04)#	0.486
Valeric acid (C5n)	0 (0–0.01)#	0 (0–0.0009)#	0 (0−0)#	0.716
Isovaleric acid (C5i)	0.001 (0–0.08)#	0 (0–0.1)#	0.01 (0–0.07)#	0.926
Dietary habits Food frequency intake assessed on a scale 1–6: 1—never or almost never; 2—once a month; 3—several times a month; 4—several times a week; 5—every day; 6—several times a day.
Sweets and snacks	2.72 ± 0.73$	2.57 ± 0.69$	2.85 ± 0.75$	0.145
Diary and eggs	3.25 (2.79–3.67)#	2.92 (2.62–3.33)#	3.5 (2.96–3.67)#	0.038 *
Cereal products	3.2 (2.8–3.6)#	3.2 (2.75–3.4)#	3.2 (3–3.6)#	0.301
Oils	2.64 ± 0.66$	2.6 ± 0.68$	2.67 ± 0.65$	0.706
Fruits	2.7 (2.48–3.12)#	2.7 (2.5–3.1)#	2.75 (2.4–3.23)#	0.629
Vegetables and seeds	3.44 ± 0.61$	3.36 ± 0.52$	3.5 ± 0.68$	0.391
Meat	2.34 ± 0.67$	2.37 ± 0.65$	2.3 ± 0.703$	0.711
Drinks (excluding water)	2.04 ± 0.55$	2.05 ± 0.59$	2.03 ± 0.53$	0.891
Processed food products	2.43 ± 0.45$	2.36 ± 0.49$	2.5 ± 0.41$	0.244
Psychometric parameters
MADRS score total	19 (16−23)#	19.5 (16−25)#	18 (14−20)#	0.079
MADRS score domain Sadness	4 (3−5)#	5 (3.5–5)#	4 (3−5)#	0.434
Neurovegetative	5.29 ± 2.03$	6.15 ± 2.16$	4.56 ± 1.61$	0.002 *
Detachment	6.83 ± 2.06$	7 ± 2.22$	6.69 ± 1.94$	0.566
Negative thoughts	3 (2−4)#	3 (2−3)#	3 (2−4)#	0.297
DASS score	62.6 ± 20.9$	62.4 ± 21.3$	62.7 ± 20.9$	0.962
Depression	20.4 ± 9.84$	19.6 ± 11.2$	21.1 ± 8.64$	0.573
Anxiety	16 (10−23)#	20 (10−24)#	14 (10−23)#	0.503
Stress	25.6 ± 8.43$	25.3 ± 7.92$	25.8 ± 8.98$	0.822
QoL score	73 ± 11.8$	74.6 ± 12.4$	71.6 ± 11.3$	0.322
Physical	18.9 ± 4$	18.7 ± 4.11$	19.2 ± 3.96$	0.648
Psychological	15.4 ± 3.42$	15.6 ± 3.6$	15.2 ± 3.29$	0.719
Social	8.5 (7−10)#	8 (7−11)#	9 (7–9.25)#	0.527
Environment	24.9 ± 4.73$	26.2 ± 4.82$	23.7 ± 4.4$	0.040 *
Circulating intestine- and microbiota-related parameters
I-FABP	1784 (1196–2305)	1920 (1217–2354)	1629 (1145–2256)	0.561
citrulline	47.9 (34−56)	48.7 (35.6–56.7)	46.6 (33.6–55.5)	0.551
bSCFAs	4672 (3535–6230)	5318 (3292–6496)	4541 (3610–5396)	0.22
propionic acid	66.4 (40.2–96.8)	64 (41.4–79.5)	67.4 (40.2–127)	0.477

Legend: $: mean ± standard deviation; #: median (interquartile range); * : p < 0.05; 6A70: single episode depressive disorder; 6A71: recurrent depressive disorder; 6A73: mixed depressive and anxiety disorder.

Abbreviations: BMI: body mass index; bSCFAS – blood short-chain fatty acids; DASS – Depression, Anxiety, and Stress Scale; F: females; I-FABP – intestinal fatty acid-binding protein; M: males; QoL – quality of life; MADRS: Montgomery–Åsberg Depression Rating Scale.

The most often used antidepressants were SSRIs, e.g., sertraline (n = 14), or escitalopram (n = 8), SNRIs, e.g. duloxetine (n = 5), or venlafaxine (n = 4), and trazodone (n = 5).

3.2. Change in fecal short-chain fatty acid levels after the intervention

No significant changes in fSCFAs were noted between the PLC and the PRO groups, in either ratios or raw values of fSCFAS (Tabl. 3). The effect size of all the changes was shown to be none or small.

A modified Intention-To-Treat (mITT) analysis was performed for the primary outcome measure for all completers (N = 64). The results were comparable to the PP analysis (p = .831, |r|=.03). Table 3

Table 3.

The change in fecal short-chain fatty acids (fSCFAs) after the intervention was expressed as ratios to total SCFAs or raw values.

	V1 PRO (median (IQR))	V2 PRO (median (IQR))	∆ PRO (median (IQR))	V1 PLC (median (IQR))	V2 PLC (median (IQR))	∆ PLC (median (IQR))	p ∆	Difference of median (Δ PRO- Δ PLC) (median [95 % CI])	Wilcoxon effect size (r)
Ratio (%)
Isovaleric acid (C5i) *	0 (0–1.91)	0.05 (0–0.65)	0 (−0.85–0)	0.14 (0–1.47)	0 (0–0.82)	0 (−0.76–0)	.853	0 [−0.12; 0]	0.02
Acetic acid (C2)	66.4 (60.9–75.4)	66.0 (62.3–77.2)	−0.91 (−8.92–6.81)	66.0 (63.9–71.8)	69.2 (65.3–75.1)	4.82 (−4.88–8.61)	.240	−5.73 [−2.6; 5.4]	0.15
Propionic acid (C3)	0.47 (0–1.91)	0.65 (0.2–1.28)	0 (−0.32–0.31)	0.77 (0.33–1.49)	0.61 (0.12–1.34)	−0.02 (−0.71–0.23)	.553	0.02 [−0.27; 0.17]	0.08
Butyric acid (C4n)	13.4 (9.32–17)	13.7 (9.81–18.3)	2.08 (−3.55–8.62)	12.8 (8.46–16.1)	12.1 (8.83–15.6)	1.47 (−4.89–4.25)	.331	0.61 [−3.1; 3.3]	0.13
Isobutyric acid (C4i)	14.4 (12.5–17.9)	14.7 (11.5–18.3)	−0.47 (−2.7–2.87)	16.9 (13.6–21.2)	16.3 (11.7–18.7)	−0.02 (−5.21–1.82)	.433	−0.45 [−2.5; 0.62]	0.1
Valeric acid (C5n)	0 (0–0.02)	0 (0–0.57)	0 (0–0.13)	0 (0−0)	0 (0–0.01)	0 (0; 0)	.412	0 [0; 0]	0.1
Raw values (mM)
Total	4.03 (3.38–6.79)	4.66 (3.01–7.28)	0.47 (−1.59–2.66)	4.54 (3.35–6.98)	4.33 (2.63–6.11)	−0.37 (−2.26–0.82)	.513	0.84 [−0.83; 0.82]	0.09
Isovaleric acid (C5i)	0 (0–0.1)	0.003 (0–0.05)	0 (−0.007–0.04)	0.01 (0–0.07)	0 (0–0.05)	0 (−0.05–0)	.150	0 [0; 0]	0.19
Acetic acid (C2)	2.9 (2.21–4.73)	3.27 (1.98–4.41)	0.33 (−1.1–1.32)	3 (2.27–4.25)	2.9 (1.97–4.05)	−0.26 (−1.2–0.82)	.551	0.59 [−0.8; 0.69]	0.08
Propionic acid (C3)	0.02 (0–0.08)	0.04 (0.005–0.08)	0.003 (−0.01–0.05)	0.04 (0.02–0.08)	0.03 (0.008–0.06)	−0.0005 (−0.03–0.01)	.088	0.0035 [−0.0095; 0.0092]	0.22
Butyric acid (C4n)	0.64 (0.27–1.21)	0.69 (0.27–1.36)	0.1 (−0.49–0.31)	0.41 (0.31–0.83)	0.51 (0.25–0.93)	0.01 (−0.37–0.25)	.580	0.09 [−0.12; 0.19]	0.07
Isobutyric acid (C4i)	0.64 (0.49–1.21)	0.72 (0.39–1.2)	−0.08 (−0.19–0.32)	0.82 (0.53–1.04)	0.77 (0.39–1.02)	−0.05 (−0.49–0.17)	.532	−0.03 [−0.17; 0.12]	0.08
Valeric acid (C5n)	0 (0–0.0009)	0 (0–0.05)	0 (0–0.01)	0 (0−0)	0 (0–0.0008)	0 (0−0)	.540	0 [0; 0]	0.08

Legend: * - primary outcome measure.

3.3. Stratification by use of antidepressants

To explore the effect on fSCFA levels of combining the intervention with antidepressant use, a subgroup analysis was performed using the Kruskal-Wallis rank sum (KW) test. The comparisons were stratified by antidepressants treatment (yes/no) and then the types of antidepressants (none / SSRIs / non-SSRIs). The PRO and PLC groups differed with regard to the observed changes of isovaleric acid (C5i) ratios (Tabl. 4 A). Pairwise comparisons using Dunn's test were used to assess differences between the PRO and PLC subgroups, as well as between the PRO subgroups (three or five post hoc comparisons).

Specifically, R0052/R0175 decreased the levels of C5i% compared with placebo when administered with antidepressants other than SSRIs (non-SSRIs) with a large effect size (p = .019, |r|=.653); no such decrease was observed when combined with SSRIs (p = .572, |r|=.109) or applied alone (p = .404, |r|=.182). Moreover, R0052/R0175 with non-SSRI antidepressants decreased C5i% levels compared with R0052/R0175 alone with a large effect size (p < .001, |r|=.810) (Fig. 2B, Table 4B). Generally, no significant differences were found when stratified by antidepressant use; however, the effect size of the analysis was found to be moderate (p = .073, η²=.07) (Fig. 2A). In post hoc analysis, R0052/R0175 with antidepressants decreased the levels of C5i% compared to R0052/R0175 alone with a moderate effect size (p = .012, |r|=.474).

Fig. 2.

The changes in isovaleric acid ratio (∆ C5i%) after the intervention with PRO vs. PLC. a. when stratified by antidepressants use; b. when stratified by the type of antidepressant treatment (none/non-SSRIs/SSRIs); c. when stratified by psychotropic mono- vs. polytherapy; * - p < .05, * * - p < .01.

Table 4.

The comparison of median change A. in fecal short-chain fatty acid ratios (Δ%) and raw values (∆); B. in fecal isovaleric acid ratio (Δ C5i%) between the probiotic (PRO) and placebo (PLC) groups stratified by particular antidepressant use.

A		Ratio									Raw values
		Δ isovaleric acid (C5i)%		Δ acetic acid (C2) %	Δ propionic acid (C3) %		Δ butyric acid (C4n) %	Δ isobutyric acid (C4i)%		Δ valeric acid (C5n) %	Δ isovaleric acid (C5i)		Δ acetic acid (C2)	Δ propionic acid (C3)		Δ butyric acid (C4n)	Δ isobutyric acid (C4i)		Δ valeric acid (C5n)
pantidepressants yes/no		.073		.524	.292		.630	.740		.507	.651		.193	.452		.154	.689		.240
p antidepressants SSRIs/non-SSRIs/no		.027		.540	.097		.608	.793		.789	.885		.243	.744		.150	.833		.128
B	SSRIs					Non-SSRIs antidepressants						No antidepressants						p total
	PRO (n = 11)		PLC (n = 16)			PRO (n = 5)			PLC (n = 8)			PRO (n = 12)			PLC (n = 9)
Δ isovaleric acid (C5i)%Mdn (IQR)	0 (−18.6 *10−2 – 0)		−9.0 * 10−2 (−83.0 *10−2 – 0)			−151.0 * 10−2 (−152.0 *10−2 – −138.0 *10−2)			0 (−3.0 *10−2 - 0)			0 (0 – 56.1 *10−2)			0 (−76.3 *10−2 – 72.9 *10−2)			.027
p	.572					.019						.404
|r|	.109					.653						.182

Legend: |r| - the absolute value of the effect size measure for Dunn’s test.

Abbreviations: SSRIs – selective serotonin reuptake inhibitors.

The changes in C5i% differed between no vs. mono- vs. poly-psychopharmacotherapy (p = .046; KW test). More specifically, mono-psychopharmacotherapy decreased C5i% to a greater degree than no psychopharmacotherapy (p = .015, |r|=.37). When stratification by PRO vs. PLC was added (six subgroups analysis), the differences were non-significant (p = .119, |r|=.068) (Fig. 2C).

No correlation was observed between fluoxetine-dose equivalents and changes in C5i%, either generally (n = 32, r = .02, p = .903) or in the PLC group (n = 20, r = -.04, p = .855). A small positive, but non-significant correlation was found in the PRO group (n = 12, r = .19, p = .564).

Abbreviations: PLC – placebo, PRO – probiotics, SSRI – selective serotonin reuptake inhibitors.

3.4. The associations between the changes in depression scores and the changes in fecal short-chain fatty acids

A correlation analysis was performed, comparing the changes in fecal SCFA ratio and MADRS score, for the PRO and separately for the PLC group.

Several differences in the correlation direction and strength of C5i were revealed between the PRO and PLC groups. Most prominently, the subjects using non-SSRIs in the PRO group demonstrated a moderately positive association between a decrease in C5i% and an improvement in depression score; however, the opposite was found in the PLC group.

Generally, in both the PRO and PLC groups, an increase in acetic acid and a decrease in propionic acid were associated with an improvement in depression score. Furthermore, among participants not receiving antidepressants, increases in butyric and isobutyric acids were positively correlated with the worsening of depression score in the PLC, but not the PRO group. In contrast, in patients not treated with antidepressants, the increase in valeric acid was positively correlated with the worsening of MADRS score in the PRO, but not PLC group (Tabl. 5).

In the parent Pro-demet study, R0052/R0175 use was shown to be associated with a higher rate of minimal clinically important difference (MCID) in MADRS score than placebo [30]. Therefore, the present analysis was additionally stratified according to the change in C5i% by MCID achievement. However, there was no difference in the change of C5i% based on MCID presence overall (p = .826) or when stratified by PRO vs. PLC intervention (p = .252). Table 5

Table 5.

Correlation analysis of the changes in fSCFAs levels and percentage changes in the MADRS score depending on the type of antidepressant treatment. * p < .05; * * p < .01.

Abbreviations: A-DASS – Anxiety-DASS; DASS – Depression, Anxiety, Stress Scale; D-DASS – Depression-DASS; PLC – placebo; PRO – probiotic; MADRS – Montgomery-Åsberg Depression Rating Scale; S-DASS – Stress-DASS; %Δ – percentage change between the end and the beginning of the intervention.

4. Discussion

Our findings indicate that fecal isovaleric acid (displayed as ratio to total fSCFAs) levels were decreased by R0052/R0175 compared to placebo, but only when combined with non-SSRI antidepressants. In addition, the allocation of R0052/R0175 or placebo was found to influence the degree of associations between the changes in specific fSCFAs and MADRS.

As fecal and cecal microbiota have been found to demonstrate similar functions, the present study discusses fecal and cecal SCFAs collectively as fecal SCFAs [41].

4.1. The changes in fSCFAs levels following the intervention

No changes in raw SCFA values or ratios, neither total nor specific acids, were found between the PRO and PLC groups after the intervention.

However, a recent meta-analysis found that pre- and probiotic treatments increased butyrate levels and improved depression scores, and that total SCFA levels seemed to be positively connected to probiotic supplementation [26]. A study on 60 healthy medical students under chronic stress revealed that heat-inactivated Lactobacillus gasseri CP2305 intake significantly reduced anxiety and sleep disturbance relative to placebo. However, this treatment had no significant effect on the raw values of SCFAs in faeces, although valeric acid levels were elevated compared with placebo, albeit with a small effect size[42]. Elsewhere, the consumption of Lactobacillus rhamnosus and additional citrus using fermented salami in 24 healthy individuals led to a significant increase in faecal butyrate. In addition, the inflammatory markers CRP and TNF-α decreased significantly after intervention, suggesting a less inflammatory environment. Treatment was also associated with improved antioxidant markers in plasma. However, gut microbiota community structure was not significantly shaped by consumption of either the regular or reformulated salami [43]. These previous studies, included in the meta-analysis, differ from the present one with regard to the duration of the intervention, which ranged from three days to four weeks, and the fact that the examined population was healthy adults. Additionally, they used bacterial strains other than R0052/R0175. It has also been shown that the levels of faecal SCFAs are influenced by baseline health-related data, including the presence of depression[44], [45], and that probiotic supplementation results in different outcomes when used by subjects with probable functional dysbiosis [46]. This may account for the differences observed between the present findings and those of our previous study. Furthermore, the meta-analysis included studies on both plasma and faeces, while our trial assessed specifically fecal SCFAs.

Several in vitro investigations have also assessed the influence of probiotic applications on the production of SCFAs. In contrast to our present findings, pretreatment with Bacillus coagulans GBI-30 followed by prebiotic resulted in increased levels of butyric, acetic, and propionic acids[47].; this difference may be accounted for the fact that different strains were used, and by the use of prebiotics as a source for microbial SCFA production [26]. Another study on in vitro gut model found that three-week supplementation with Lactobacillus and Enterococcus species, using a stabilized microbial community derived from the faecal microbiota of healthy human donors, led to increased production of total SCFAs[48]. Additionally, the probiotics were found to demonstrate an immunomodulatory effect, characterised by increased production of anti-inflammatory cytokines (IL-10 and IL-6) and lowered production of inflammatory chemokines (IL-8, CXCL10, and MCP-1). However, unlike the present study, the intervention targeted the faecal microbiota of healthy subjects and not that of a clinical population with depression; it also used a different probiotic strain.

Interestingly, it was shown that the pre-treatment levels of faecal SCFAs influence the degree to which probiotics alter their production. In particular, Lactobacillus paracasei reduced faecal butyrate concentration by nearly 50 % in butyrate-rich faeces (>100 mmol/kg) compared to threefold in butyrate-poor stool (<25 mmol/kg). These changes occurred together with shifts in some bacterial genera [46]. Thus, fecal SCFAs levels may serve as biomarkers for identifying species that could gain an advantage from probiotic intervention; however, in the present study, controlling for baseline levels of acetic, butyrate, and isobutyrate acids did not confirm this hypothesis. In addition, no such analysis was possible with C5i as a primary outcome measure, as this was absent from a high number of samples. This is to be expected as C5i represents an exceedingly small amount of total SCFAs [49], [50].

It is possible that R0052/R0175 did not influence the fecal levels of SCFAs due to several factors, particularly the heterogeneity of our sample with regard to medications used, comorbidities, and metabolic status; all are known to influence the status of the basal microbiota, and thus the action of probiotics. While subpopulation analyses could reveal distinct results, they would be highly underpowered due to the large number of subgroups and a small number of participants per subgroup.

4.2. The influence of probiotic application and antidepressant treatments on changes in fSCFA levels

We found that R0052/R0175 as an add-on to non-SSRIs decreased the levels of C5i% compared with placebo plus non-SSRIs or R0052/R0175 alone. Moreover, our study provides preliminary data on the increase in fecal isovaleric acid levels after the intervention with R0052/R0175 as an add-on to non-SSRIs compared with add-on to SSRIs or R0052/R0175 alone. In the non-SSRIs group, the most commonly used drugs were SNRIs or trazodone. All of these antidepressants inhibit not only serotonin (5-HT) reuptake, but they also act as norepinephrine (NE) reuptake inhibitors or modulators of several receptors [51], [52].

4.2.1. Dual action on serotonin and norepinephrine transmission

It was already shown that probiotics increased the turnover of 5-HT, NE, or dopamine (DA) in the prefrontal cortex or hippocampus in animal models of depression [53], [54], [55]. In addition, probiotics have been found to change the balance of different SCFAs inconsistently under specific conditions [26], [42], [46], [47], [48], [55].

For example, regarding the influence of antidepressants on SCFA levels, fluoxetine was found to decrease the levels of C4i and increase C4n in a rat model of depression [56]., while different classes of antidepressants appeared to increase the faecal levels of C2, C3, C4n, and C5i after treatment in animal models of depression [57]; however, no subgroup analyses were conducted. Moreover, berberine was shown to alter the levels of SCFAs, including isovaleric acid, and increase the turnover of 5-HT, NE, and DA in the hippocampus in animal models of depression [58]. These findings highlight the significance of the relationship between monoamine expression and changes in SCFA levels in depression.

Therefore, we hypothesize that interventions aimed at influencing both 5-HT and NE turnover might be more effective at changing SCFA levels than targeting mainly serotoninergic transmission. Moreover, intervention with probiotics together with non-SSRIs may have additive or even synergistic effects. As C5i is known to play a key role in depression [21], [22], [23], [24], it may also represent a significant target for interventions, as shown in our primary outcome measure analyses.

4.2.2. Alleviated dysbiosis as a possible explanation

Antidepressants such as SSRIs or SNRIs can exert direct antibacterial effects on gut microbes, and indirectly alter the microbial environments via their pharmacodynamic effects on the host [59], [60], [61]. Notably, sertraline, fluoxetine, paroxetine, and escitalopram (all SSRIs) exhibit significant antibacterial properties [62]. Previous studies have revealed a correlation between gut microbiota diversity and various antidepressants, including SSRIs/SNRIs, in humans [29], [63], [64], [65]. For example, Gao et al. found that patients with depression have a distinct gut microbiome that changes after SSRIs treatment [65]. Similarly, another study reported that escitalopram improved the Firmicutes/Bacteroidetes ratio and alpha diversity, although some structural and metabolic differences persisted. These residual disparities in the microbiota may be linked to depression relapse, highlighting the complex relationship between gut health and antidepressant efficacy [64]. More specifically, Turicibacter was found to be depleted by SSRI medications, with a bacterial protein resembling a serotonin transporter being proposed as their target [29], [66]. Importantly, those changes were repeatedly shown to develop gradually and to result from long-term use of antidepressants [29]. A number of animal studies have examined the impact of various classes of antidepressants on microbiota taxonomy or metabolites [67], [68].

Probiotics like Lactobacillus and Bifidobacterium help shift the gut environment toward more carbohydrate fermentation (via fibre), which promotes the production of beneficial SCFAs like butyrate, while reducing protein fermentation and isovaleric acid production [69], [70]. Thus, R0052/R0175 may help mitigate SSRI/SNRI-induced microbial imbalances, which may reduce byproducts like isovaleric acid. However, R0052/R0175

as an add-on to SSRIs may not be enough to restore microbiota balance, but they may be enough with non-SSRI antidepressants. The explanation may lie in possible variation in microbiota alterations caused by SSRIs vs. non-SSRIs. As discussed above, SNRIs are probably less likely to affect microbiota directly, though the data is much less robust.

4.2.3. Changes in intestinal motility as an interplaying factor

Moreover, different influences of R0052/R0175 with SSRIs or non-SSRIs on fecal isovaleric acid (C5i) levels may be further due to the antidepressants' distinct impact on gastrointestinal motility. While SSRIs’ common side effects include diarrhea, escitalopram and sertraline were shown to affect the gastrointestinal tract the most severely [71], [72]. At the same time, these two antidepressants were the most commonly used SSRIs by the subjects of our study. SNRIs, on the other hand, may cause constipation in up to 15 % of their users, as norepinephrine may slow intestinal transit, especially through its influence on sympathetic nervous activity [71], [73]. On the other hand, trazodone, the most robustly used non-SSRI in our study, has been connected to constipation through both anticholinergic action and slowed intestinal motility [74], [75]. The longer the bowel content is present in the colon, the more microbiota metabolites may be absorbed into the circulation. Essentially, the intake of probiotics was shown to be associated with better absorption of many nutrients [76] This may further support our findings on decreased fecal C5i in patients supplemented with R0052/R0175 and using non-SSRI antidepressants.

4.2.4. Additional factors

Or findings indicate that psychotropic mono-psychopharmacotherapy was associated with reduced levels of C5i% compared with no psychopharmacotherapy; however, no difference was observed between poly-psychopharmacotherapy and no psychopharmacotherapy. Additional stratification by PRO vs. PLC resulted in non-significant differences; however, this was probably due to the low numbers of subjects per group when six subgroups were analyzed. Indeed, a prior study linked general polypharmacotherapy to decreased microbiota diversity [77]. Thus, the interaction between polypharmacy and microbes might result in unfavorable changes in SCFA production. By analogy, in the present study, treatment with a single antidepressant may be connected with lower concentrations of C5i, which have been associated with depression [21]. Using more than one psychotropic drug was not found to have the same effect. However, due to preliminary character of our study, further investigation is needed to uncover the influence of mono- vs. poly-psychotropic medication on microbiota metabolite levels, and determined to role of probiotics in this interaction.

Additionally, the gut microbiota directly influences drug metabolism. Bacterial enzymes can metabolize compounds such as tryptophan, essential for serotonin production, while some microbes can influence the availability of drugs like duloxetine through bioaccumulation. Furthermore, microbial activity can modulate hepatic enzyme expression or gene activity, altering how the host processes medications. Microbial metabolites, such as SCFAs and secondary bile acids, act as potential mediators in these processes, influencing both drug efficacy and host responses [78]. This bidirectional interaction between antidepressants and gut microbiota highlights the importance of considering microbial health in designing effective pharmacological therapies.

4.3. The relationship between change in fSCFA content and change in depression score

The main finding of the parent Pro-demet study was that probiotics minimally, but significantly, decreased depression severity score when administered as an add-on; however, not suh change was noted when given as standalone treatment. Moreover, the action of probiotics was dependent on pre-intervention metabolic health status [30]. By analysing the changes in depression score and fSCFA levels in different intervention groups, the present study revealed several interesting trends for all participants, and for the PRO or PLC plus antidepressant intervention subgroups.

A very recent study by Jiang et al. examined the relationship between gut microbiota composition and the efficacy of SSRIs. Significant differences in microbiota were observed between responders and non-responders, with taxa such as Ruminococcus, Bifidobacterium, and Faecalibacterium being more abundant in the responder group. Functional analysis revealed upregulation of acetate degradation and neurotransmitter synthesis pathways in the responder group. The machine learning model indicated that gut microbiota and metabolites are potential biomarkers for predicting the efficacy of SSRIs [79].

Less is known of the changes of faecal SCFAs after antidepressant treatment. In our study, in the absence of any intervention, an increased depression score was associated with an increase in butyrate and isobutyrate levels. It is possible that this is a compensatory natural mechanism, as endogenous butyrate and its derivatives have been repeatedly shown to be associated with lower depression severity in animals and humans [26], [55]. For instance, Muller et al. found depressive symptoms to correlate positively with acetate levels but negatively with levels of butyrate and propionate [80]. Similarly, Waseem et al. indicate that anxiety symptoms were correlated negatively with butyrate levels but positively with acetate levels [81]. Also, it has been suggested that butyrate and other SCFAs may improve depression-like behaviors through elevating brain-derived neurotrophic factor (BDNF) levels, which led to the use of butyrate and propionate as an experimental drug in models of neuroinflammatory disorders, including depression [82]. However, no such association was observed in the PRO group of our present study, indicating that probiotics may exert their antidepressant action by more complex mechanisms than only through C4 or C4i. Indeed, a recent meta-analysis found that while butyrate and total faecal SCFA concentrations increased, linear regression models did not reveal no significant direct correlation between SCFAs and depression [26].

Interestingly, our correlation analysis found an improvement in depression was related to a decrease in C5i% in subjects using non-SSRIs in the PRO group, but not in the PLC group. Importantly, Szczesniak et al. reported that fecal isovaleric acid was the only SCFA to display a direct, significant correlation with depression, with depressed patients being overrepresented in the group with higher isovaleric acid levels (p < .00005). A positive correlation was also found between C5i and cortisol, as well as between C5i and microbes previously identified to be associated with depression [21]. Similarly, a study reported decreased anxiety and depression behaviours in patients with irritable bowel syndrome comorbid with symptoms of depression and anxiety who underwent faecal microbiota transplantation (FMT); these changes were accompanied by decreases in the levels of faecal C5n and C5i compared with placebo. Moreover, FMT lowered the abundance of Faecalibacterium, Eubacterium, or Escherichia [83]. In another study, Jiaotaiwan, a classic traditional Chinese medicine prescription, was shown to improve depression as an add-on supplement compared with an antidepressant alone. At the same time, the combination increased the serum levels of C5i and C4i, along with brain derived neurotrophic factor (BDNF) levels [84].

As regards animal studies, Lactiplantibacillus plantarum R6–3 was found to exert a higher antidepressant effect in a mouse model of depression compared to fluoxetine. Simultaneously, this probiotic increased the secretion of C4n, C5n and C5i by nearly 50 % compared with antidepressant [85]. It is contradictory to our study findings, but apart from the nature of the study (a mouse model), a different probiotic strain was used. Additionally, berberine, a compound known for its antidepressant effects, has also been shown to alleviate depression-like behaviors in depressed rats by modulating the gut microbiota and increasing fecal SCFAs concentrations, such as C2, C3, and C5i acids. Further, berberine increased the expression of key neurotransmitters, including serotonin, norepinephrine, dopamine, as well as brain-derived neurotrophic factor in the hippocampus, highlighting its potential to influence the microbiota-gut-brain axis [58]. The difference with our study results may be since not a probiotic, but different antidepressant agent was used. Finally, the time of the antidepressant intervention in the above animal studies was much shorter than in our study; thus, the bi-phasic action of the intervention on fC5i levels is possible.

Summarizing, the differences between the results of our study and the above-referenced studies may arise from several factors. First, some of the studies were conducted on animal models of depression. Second, the intervention period varied between several days and two months. Third, the pre-intervention microbiota composition and function may play significant roles in the mechanisms through which treatment response is achieved or not. Then, the assessment of depression in humans was performed with different methods, including professional assessment or self-assessment scales. Last but not least, different methods were applied to evaluate SCFAs levels. All in all, C5i seems to be a significant factor in the interplay between antidepressant efficacy and microbiota.Even though probiotics in the parent study were shown to result in minimal improvement in depression score when given as an add-on treatment, no differences in C5i% were revealed when the analysis was stratified by MCID and by PRO vs. PLC allocation. However, different changes in microbiota diversity and specific composition have been noted in SSRI responders compared with non-responders [86]. This discrepancy may again be due to the very small number of subjects in our analysis subgroups, and the fact that the analysis stratified by PRO/PLC and antidepressant use lacked sufficient power to be performed in the present study. To sum up, we assume that the decrease in fecal C5i may contribute to the antidepressant efficacy of probiotics as an add-on to non-SSRIs intervention in human subjects after long-term treatment.

4.3.1. Isovaleric acid (C5i) possible antidepressant action

It may be hypothesized that the lower fecal levels of C5i reflect the higher circulating and brain levels of this metabolite in patients treated with non-SSRIs combined with R0052/R0175 [87]. As shown, the decrease in fecal isovaleric acid was connected to the improvement in depression in this group of patients. This might be since C5i may act peripherally and centrally through HPA or brain-derived neurotrophic factor (BDNF) thus enhancing antidepressant activity [21], [84]. Indeed, microbiota-targeted antidepressant therapies were previously shown to be connected to BDNF signaling [88] or normalize HPA activity [89]. Moreover, C5i may possess an antianxiety action thanks to modulation of GABA-A receptors [90]. As depression is highly comorbid with anxiety at different levels [91], [92], counteracting anxiety may strongly influence the antidepressant action of the intervention.

Essentially, the data on simultaneous changes in depression severity and fecal SCFAs is scarce, and our study may background for future research.

4.4. Limitations

The study included a small sample size, which allows for only a preliminary analysis of the results; as such, more studies with larger sample sizes are needed to confirm the findings. In addition, no data on the duration of antidepressant use before the intervention was available. Further, the study did not assess circulating SCFA levels. The results also have limited generalizability,as the study group was mostly composed of Caucasian women, and the data is closely associated with the particular probiotic strain used in the formulation.

5. Conclusions

The use of probiotics as an add-on may improve the antidepressant action of non-SSRI antidepressants, partly through their interaction with isovaleric acid (C5i). These findings justify further investigations assessing the specific conditions for psychobiotics application in the field of mood disorders.

Informed consent statement

Informed consent was obtained from all the subjects involved in the study.

Funding

This research was funded by the Medical University of Lodz, grant numbers 503/1–155–02/503–11–003–20, 502–03/1–155–02/502–14–386–18, and 564/1–000–00/564–20–070.

Institutional review board statement

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Ethics Committee of the Medical University of Lodz (15th December 2020; reference number RNN/228/20/KE).

Declaration of Competing Interest

K.S.-Ż. receives remuneration from a probiotic company. Other authors have nothing to declare.

Appendix A. Supplementary material

Supplementary material

Data Availability

The data that support the findings of this study are available from the corresponding author, O.G.-K., upon reasonable request.