Daily consumption of Fuzhuan brick tea improves lipid profile and vascular function associated with modulation in gut microbiota composition in healthy adults: a pilot randomized controlled trial

Abstract

Background

Dietary strategies play a pivotal role in the prevention of cardiometabolic diseases. Animal studies have shown that Fuzhuan brick tea (FBT), a popular post-fermented tea, enhances cardiometabolic health by modulating the gut microbiota; however, there is no definitive evidence from human trials. This 12-week randomized controlled trial investigated the effects of FBT on the lipid profile, vascular function, and gut microbiota in healthy adults.

Methods

This 12-week, open-label, parallel-group, randomized controlled trial was conducted in a rural community in northern China from September to November 2021. After giving informed consent, 120 healthy adults recruited were randomly assigned to either a group consuming two daily glasses of brewed FBT infusion (totaling 12 g/day) or a group consuming an equivalent amount of water. Serum lipid levels and vascular function, as indicated by flow-mediated dilation and nitroglycerin-mediated dilation, were measured at baseline (0 weeks) and 12 weeks after intervention. Fecal samples collected at baseline, 6 weeks, and 12 weeks underwent profiling of microbiota composition by shotgun metagenomic sequencing. This trial was completed and registered at chictr.org.cn (registration number ChiCTR2100047294).

Results

Compared to water, consumption of FBT resulted in significantly reduced low-density lipoprotein cholesterol levels, with improved vascular function (the co-primary outcomes). The longitudinal metagenomic analysis detected no significant alterations in richness and distribution (alpha diversity). However, a subtle but statistically detectable shift in the gut microbiota composition (beta diversity) was observed. Exploratory species-level analyses suggested increases in bacteria that produce short-chain fatty acids, such as Roseburia hominis, Eubacterium saphenum, Eubacterium sp. AF15 50, Alistipes dispar, Blautia stercoris, and Lactobacillus amylolyticus, which were correlated with the lipid profile or vascular function.

Conclusion

Daily consumption of FBT demonstrated beneficial effects on lipid levels and vascular function in healthy adults, and was associated with exploratory shifts in gut microbiota composition, supporting FBT consumption is a simple, affordable, and practical dietary strategy for improving cardiometabolic health.

Clinical trial registration

http://www.chictr.org.cn/, identifier ChiCTR2100047294.

Introduction

Cardiometabolic diseases (CMDs), including diabetes, insulin resistance, obesity, stroke, nonalcoholic fatty liver disease, and cardiovascular disease, remain the primary cause of morbidity and mortality globally among all chronic non-communicable diseases (1, 2). Recognizing the urgency of addressing CMDs, global health authorities have prioritized this burden and developed policies aimed at ameliorating cardiometabolic risk factors. Clinical guidelines underscore dietary strategies as fundamental in preventing and managing CMDs (3). While many functional foods show promise in preventing and treating CMDs, the diversity in dietary habits across different countries and regions presents a considerable challenge in identifying functional foods that are compatible with local dietary practices.

Tea (Camellia sinensis), one of the most widely consumed beverages worldwide, offers a range of bioactive compounds that provide antioxidant, anti-inflammatory, and lipid-lowering effects (4–6). Consequently, tea has been identified as a promising functional food with potential benefits for CMDs prevention. Chinese tea can be classified into six major types based on their manufacturing processes and fermentation levels: non-fermented green tea, white tea, yellow tea, partially fermented oolong tea, fully fermented black tea, and post-fermented dark tea (7). Dark tea, including Pu-erh tea and Fuzhuan brick tea (FBT), refers to tea that undergoes microbial fermentation (8). FBT, also known as “Hu Tea” or “Jinyang Zhuan,” originated from Jingyang, Shaanxi Province, China, during the Ming Dynasty in the 16th century and is now also produced in Hunan and Guizhou provinces (9, 10). It has been a staple beverage among ethnic groups residing in the border regions of northern and western China. Notably, FBT undergoes a distinct “fungal fermentation” stage during its manufacturing process. Controlled temperature and moisture conditions facilitate the growth of yellow fungi, known as “golden flower,” within the tea. These fungi consist of a mixture of various microorganisms, with Eurotium cristatum being the dominant species (9, 11, 12). FBT also possesses unique functional ingredients, including polyphenols, polysaccharides, and soluble dietary fibers. In comparison to unfermented tea, such as green tea, FBT exhibits varying degrees of reduction in polyphenol content and higher levels of polysaccharides (10). In recent years, accumulating evidence in animal studies has shown that FBT and its bioactive compounds possess a diverse range of biological activities, such as antioxidant, anti-inflammatory, hypolipidemic, anti-obesity, anti-bacterial, and anti-tumor activities (10, 13, 14). Given its widespread consumption in northwest China, FBT consumption holds promise as a simple, affordable, effective, and practical approach to enhancing cardiometabolic health. However, existing evidence primarily stems from animal experiments, with limited research on humans, particularly randomized controlled trials.

With the growing recognition of tea’s diverse health benefits, researchers have begun to focus on the crucial involvement of the gut microbiota in these processes. The human gut microbiota, a complex ecosystem hosting an astonishing 100 trillion microbes across approximately 5,000 species, plays a pivotal role in host metabolism by modulating various physiological functions, including gut barrier integrity, energy uptake, and host immunity. Epidemiological and clinical studies have established connections between gut microbiota and CMDs (15–17). Mechanistic studies in animal models have further suggested a causal role of gut microbiota in the pathogenesis of CMDs through specific microbial or shared microbiota-host metabolites, identifying potential mammalian targets (15, 16). Consequently, the gut microbiome has emerged as a promising therapeutic target for CMDs. Diet significantly influences microbial abundance and functionality (18). Previous studies have demonstrated that tea consumption improves cardiometabolic health in animal models by modulating the gut microbiota (19, 20). However, studies involving humans in this area are limited and have yielded inconsistent results (21–23). Furthermore, these studies predominantly rely on the 16S rRNA gene sequencing technique. The advent of high-throughput sequencing has enabled more comprehensive analyses of the gut microbiota composition through metagenomic approaches (24, 25). Given the lack of human studies and the emergence of advanced sequencing technologies, there is a compelling need to explore the effects of tea consumption on gut microbiota using metagenomic sequencing. This approach may elucidate the relationship between changes in the gut microbiota and cardiometabolic risk factors, offering insights into potential therapeutic strategies.

In light of these considerations, we aimed to bridge the gap in human research by conducting a 12-week RCT to investigate the effects of daily FBT consumption on cardiometabolic health indicators, including lipid profile and vascular function, in healthy adults. Specifically, we employed advanced metagenomic sequencing techniques to comprehensively analyze changes in the gut microbiota composition associated with FBT consumption. Our objective was to deliver a comprehensive analysis of the changes in gut microbiota associated with FBT consumption and their correlation with cardiometabolic health indicators. This study represents a pioneering effort to explore the impacts of FBT on human health, providing valuable insights into dietary strategies for the management and prevention of CMDs.

Methods

Participant recruitment

In this RCT, all participants were recruited from healthy adults residing in a rural community in northern China in September 2021, ensuring uniform demographic characteristics, consistent environmental exposure, analogous dietary habits, and comparable physical activity patterns. The trial was registered at chictr.org.cn under registration number ChiCTR2100047294 on June 11, 2021. Candidates were invited for onsite screening interviews. Eligible subjects were healthy men and women between 20 and 65 years of age. Individuals were excluded from the study if they: (1) had a history of hypertension, diabetes mellitus, liver or renal diseases, enteritis, intestinal obstruction, inflammatory bowel disease, hypothyroidism, mental illness, stroke, or heart disease, or any chronic condition requiring long-term medication; (2) had a history of gastrointestinal tract surgery or disease; (3) had received prebiotics or probiotics, antibiotics, laxatives, anticholinergics, or antidiarrheal drugs more than once per week within 4 weeks before enrollment; (4) had received surgical treatment within 3 months; (5) were pregnant or lactating; (6) had participated in other clinical studies within 3 months before enrollment; (7) were regular tea drinkers or in the case of occasional tea drinkers, had not adhered to a 2-week washout period prior to starting the intervention; or (8) had other health issues unsuitable for participation. A total of 342 candidates were screened, and 120 of them were eligible and consented to participate in the study. The institutional ethics committee of Shaanxi Provincial People’s Hospital approved the study protocol, and each participant gave written informed consent. All procedures were performed in accordance with institutional guidelines.

Intervention strategies

The present study was an open-label, parallel-group RCT. Participants were randomly allocated to the FBT group (n = 60) or the control group (n = 60) in a 1:1 ratio. Randomization was conducted using computer-generated random numbers, employing a stratified block design stratified by sex (male/female), with a block size of 6 within each stratum. Allocation was concealed using sequentially numbered, opaque, sealed envelopes (SNOSE). The randomization sequence was generated by Yudan Zhang, who was not involved in participant enrollment or intervention assignment. Participants in the FBT group consumed 2 × 300 mL of FBT infusion per day (2 standard cups of 300-mL hot water, each containing a 6-g FBT capsule) for 12 weeks. The control group consumed an equivalent volume of hot water for the same intervention period. All participants were instructed to adhere to their regular diet and activity patterns. They were also required to complete a three-day dietary record and the International Physical Activity Questionnaire at baseline (Week 0), Week 6, and Week 12. Participant compliance was monitored through WeChat check-ins and monthly follow-up visits. Medication and supplement usage, as well as adverse events, were self-reported throughout the study.

Preparation of FBT

FBT was purchased from Xianyang Jingwei Fu Tea Co. Ltd., China, and underwent a standardized grinding process to prepare a homogenous fine powder. Then, the powder was portioned into precise 6-g single-serve tea capsules (sealed containers, not intended for oral ingestion). A dedicated brewing device was designed and calibrated to deliver a fully standardized preparation protocol: each capsule was placed into the device, which automatically dispensed a fixed volume of 300 mL water at a constant temperature of 90 °C and maintained a precise brewing duration of 5 min. This dedicated apparatus ensured uniform extraction conditions across all participants and all administration occasions throughout the 12-week intervention, thereby minimizing inter-occasion variability in bioactive compound delivery. Participants consumed only the brewed tea infusion; the spent capsule and residual tea powder were discarded. FBT contains 23.29 ± 0.09 mg/g of free amino acids (FAAs), 52.19 ± 0.51 mg/g of soluble sugars (SSs), 3.62 ± 0.03% and 32.55 ± 0.7 mg/g of caffeine (CAF), 103.18 ± 0.9 mg/g of tea polyphenols (TPs), and 16.92 ± 0.79 mg/g of total flavonoids (TFs). Additionally, it contains (16.5 ± 0.71) × 10⁵ CFU/g of Eurotium cristatum. The FFAs, CAF, and TPs were determined according to the spectrophotometry method used in our previous study. The SSs and TFs were measured using colorimetric methods based on anthrone and aluminum chloride. The TPs and TFs were quantified based on the equivalent content of gallic acid and rutin, respectively (26). The 12 g/day dose was selected based on traditional FBT consumption patterns in northwestern China (15–30 g/day) (9, 10), human equivalent doses derived from animal studies (approximately 8–15 g/day).

Specimen collection

Stool samples were collected at baseline, Week 6, and Week 12. These samples were initially stored in the participants’ home freezers at −20 °C and wrapped in ice packs before being transported on ice to the research laboratory for storage at −80 °C. Fasting blood samples were collected from an antecubital vein by a phlebotomist using standard clinical procedures at baseline and Week 12. For serum analysis, blood was collected in serum-separating tiger top tubes, centrifuged at 1,200 × g for 10 min, aliquoted, and stored at −80 °C. For plasma analysis, blood was drawn into ethylenediaminetetraacetic acid tubes, centrifuged under the same conditions, aliquoted, and similarly stored at −80 °C.

Biochemical and hematologic parameter analysis

Fresh serum samples were promptly dispatched to Shaanxi Provincial People’s Hospital on the same day of collection for the meticulous analysis of lipid-lipoprotein profiles using well-established standard techniques. The total cholesterol (TC) levels were determined by an enzymatic colorimetric method using cholesterol esterase, cholesterol oxidase, peroxidase, and chromagen 4-aminophenazone/phenol. Triglyceride (TG) levels were determined by an enzymatic colorimetric method using lipoprotein lipase glycerokinase, glycerophosphate oxidase, and chromagen 4-aminophenazone /N-ethyl-N -(3-sulphopropyl)-m-amisidine. The high-density lipoprotein cholesterol (HDL-C) levels were determined by an immunoinhibition assay. The low-density lipoprotein cholesterol (LDL-C) levels were directly measured in fasting serum using a homogeneous direct enzymatic method (hydrogen peroxide scavenging method) on a Hitachi LABOSPECT 008AS automated clinical chemistry analyzer (Hitachi High-Technologies Corporation, Tokyo, Japan), using reagents from Maccura Biotechnology (Chengdu, China), according to the manufacturer’s instructions. The investigators analyzing the samples were blinded to the identities of the samples.

Vascular function measurements

To evaluate the vascular function of participants, flow-mediated dilation (FMD) was conducted in line with established guidelines. Prior to the FMD test, participants were instructed to abstain from smoking, consuming food, or ingesting caffeine-containing beverages for 12 h. The FMD test was scheduled between 9 a.m. and 10 a.m. Participants were required to lie down and rest for a minimum of 15 min in a calm environment with a controlled temperature of 25 °C. Using a high-resolution ultrasound device (Vivid I; GE Healthcare, Trondheim, Norway), the diameter of the brachial artery was measured over a 5 cm segment. Baseline measurement was taken for 30 s, followed by measurement during peak reactive hyperemia, which occurred after the release of a blood pressure cuff that had been placed around the forearm and inflated to 50 mmHg above systolic blood pressure for 5 min. The results were quantified as the percentage increase in arterial diameter from baseline to the maximum post-cuff release. To assess endothelium-independent vasodilation, the nitroglycerin-mediated dilation (NMD) was measured. After a rest period of at least 15 min following the FMD test, the arterial diameter response was recorded following the sublingual administration of 0.5-mg nitroglycerin. The NMD was calculated based on the increase in arterial diameter measured 5 min post-nitroglycerin administration, expressed as a percentage of the baseline diameter. The investigators conducting the vascular function measurements were blinded to the group allocation.

DNA preparation and shotgun metagenomics sequencing

Metagenomic sequencing and bacterial DNA extraction were conducted using established protocols. Bacterial DNA from fecal samples was extracted using the MagBeads FastDNA™ Kit (MP Biomedicals, Shanghai, China), following the manufacturer’s instructions. The DNA concentrations were determined using NanoDrop 2000, and their integrity was assessed by 1% agarose gel electrophoresis using the KAPA HyperPlus PCR-free reagent (Illumina, CA, USA). The DNA was then sheared to an average size of approximately 400 bp for paired-end library construction. Qubit 2.0 (Thermo Fisher, Massachusetts, USA) was used to quantify the DNA samples, and Agilent 2,100 (Agilent, Palo Alto, USA) was used to verify the read quality. Libraries from all samples were pooled, ensuring a total DNA amount of 300 ng in a volume of ≤48 μL. Following heat denaturation, single-strand circularization, enzymatic digestion, and purification, shotgun metagenomic sequencing was executed on the BGI platform (MGESEQ-T7, Shenzhen, China) by Shaan Probiomicros Co., Ltd. (Xi’an, China). Raw reads underwent quality control using fastp (v0.23.0) with parameters set to length_required = 50 and n_base_limit = 2. Host reads were subsequently filtered out using Bowtie2 software (v2.3.5.1) based on human genomes.

Metagenomic data analysis

Following rigorous quality control of the sequencing data, MetaPhlAn4 software was used to annotate the metagenomic data and analyze the composition of microbial communities to obtain relative abundance information at different taxonomic levels (27). Microbiota richness and distribution were assessed using α-diversity indices, using three different indices (i.e., Shannon, inverse Simpson, and Simpson). The statistical significance of alpha diversity between the two groups at different time points was evaluated by the Mann–Whitney U test or Student’s t-test. Compositional disparities between the two groups were discerned through principal coordinates analysis (PCoA) based on the Bray–Curtis distance. PERMANOVA was performed with group as the main factor and age, gender, alcohol intake, and smoking as covariates to assess the independent effect of FBT on beta diversity. To verify that PERMANOVA results were not confounded by differences in within-group dispersion, the betadisper test was performed on the Bray-Curtis distance matrix at week 12. The influence of group categorization and other variables on the gut microbiota was evaluated using permutational multivariate analysis of variance. Longitudinal analysis was performed using SplinectomeR (28). Differences in taxon abundances over the course of the study were determined using the permuspliner function of SplinectomeR, a permutation-based package in R utilizing weighted local polynomials to assess group differences in longitudinal microbiota data, with multiple time points applied per subject.

For functional profiling, HUMAnN3 was utilized to obtain information on species abundance, metabolic pathways, and functional modules of microorganisms. The Gut Metabolic Modules (GMMs) database, including 103 metabolic pathways, was employed to analyze and annotate microbial functional changes. The GMMs database captures bacterial and archaeal metabolic pathways specific to the human intestinal environment, emphasizing anaerobic fermentation processes (29). The GMM abundance was estimated using Bowtie2 (v2.3.5.1) (30), which was constructed using data from an extensive literature review and metabolic databases, including Kyoto Encyclopedia of Genes and Genomes (KEGG) and MetaCyc. A species was deemed to possess a GMM if it harbored at least two-thirds of the necessary orthologs according to the KEGG of a module. To determine the abundance of each module, we calculated the median abundance of these KEGG orthologs within the pathway that exhibited the maximum coverage.

Statistical analysis

This study was designed as an exploratory pilot study. The sample size of 120 participants was determined based on feasibility and available resources rather than a formal sample-size calculation. The primary outcome of this study was the change in lipid–lipoprotein profiles and vascular function parameters. Secondary outcomes were changes in the gut microbiota composition after FBT intervention. All primary and secondary efficacy analyses were performed on the per-protocol population. All data were checked for normality using the Shapiro–Wilk test before further analyses. Continuous outcomes were analyzed using the t-test or Wilcoxon rank sum test, as appropriate. Dichotomous outcomes were analyzed using Fisher’s exact test. The comparison of lipid–lipoprotein profiles and vascular function parameters in the stage before and after the intervention in each group was performed using a paired sample t-test. To account for the repeated-measures design and within-subject correlations, linear mixed-effects models (LMM) were employed to evaluate the effects of FBT on the primary outcomes. For each outcome, group (FBT vs. control), time (baseline vs. week 12), and their interaction were included as fixed effects, with age, sex, alcohol intake, and smoking as covariates, and subject as a random intercept. The group × time interaction was used to assess whether FBT produced significantly different temporal changes compared with the control group. As a sensitivity analysis, analysis of covariance (ANCOVA) was additionally performed for primary outcomes, with the post-intervention value as the dependent variable, group assignment as the fixed factor, and the corresponding baseline value as the covariate. A second model further adjusted for age, sex, and BMI. Longitudinal analysis was performed using SplinectomeR. A p-value of 0.05 was deemed statistically significant. The species taxonomic level was selected for analysis. The connection between species and clinical data was assessed using Spearman’s rank correlation. Correlation analyses were performed using pooled data from baseline and week 12 to capture both pre- and post-intervention variability. All correlation p-values were adjusted for multiple testing using the Benjamini-Hochberg FDR procedure. To address the potential non-independence of pooled repeated observations, a sensitivity analysis was additionally performed using week-12 data only. Statistical analysis and visualization were carried out in the R environment V.3.4.2¹.

Results

Demographic and clinical characteristics

The CONSORT diagram and flowchart of participants through the study are depicted in Figure 1. A total of 342 individuals were initially screened for eligibility, of whom 120 were eventually enrolled. During the 12-week intervention, 10 participants discontinued (FBT group, n = 5; control group, n = 5). The reasons for discontinuation were hospitalization due to infection (n = 1), gastrointestinal discomfort unrelated to the intervention (n = 2), initiation of statin therapy (n = 3), initiation of antibiotic therapy (n = 3), and loss to follow-up (n = 1). All statistical analyses were performed on the per-protocol population, comprising the 110 participants who completed the intervention and provided valid biological samples at both baseline and week 12. The baseline characteristics, including anthropometric and blood biochemical traits, were well-balanced between the two groups (Table 1). Baseline characteristics did not differ significantly between completers and non-completers (Supplementary Table 1). With respect to potential confounders, neither three-day dietary records nor IPAQ-assessed physical activity showed significant between-group differences at any time point, nor significant within-group changes from baseline to week 12 (all p > 0.05; Supplementary Table 2), indicating that dietary habits and physical activity levels remained stable throughout the intervention in both groups.

Figure 1.

The CONSORT study flow diagram and study design. (A) Participant flowchart. (B) Study design.

Table 1.

Demographic information.

Variable	Experimental group (n = 55)	Control (n = 55)	Statistic Value	p value
Age (years), Mean ± SD	53.82 ± 8.40	53.98 ± 7.98	0.108	0.917a
Sex, n (male/ female)	31/24	32/23	0.037	0.874b
Alcohol use, n (y/n)	25/30	22/33	0.334	0.563b
Smokers, n (y/n)	16/39	19/36	0.377	0.539b
BMI, kg/m2	25.52 ± 2.95	24.45 ± 3.08	1.857	0.066a
SBP, mmHg	122.41 ± 13.31	126.02 ± 13.42	1.412	0.161a
DBP, mmHg	77.71 ± 8.16	79.54 ± 6.56	1.3	0.196a
HbA1c, %	5.83 ± 0.31	5.79 ± 0.28	0.672	0.503a
BUN, mmol/L	5.36 ± 1.44	5.38 ± 1.25	0.129	0.898a
SCr, μmol/L	58.43 ± 12.84	57.41 ± 11.48	0.438	0.663a
AST, U/L	20.85 ± 7.67	21.04 ± 6.74	0.132	0.895a
ALT, U/L	19.45 ± 10.93	18.88 ± 10.45	0.285	0.776a
Total cholesterol, mmol/L	4.82 ± 1.06	5.06 ± 0.85	1.315	0.191a
Triglycerides, mmol/L	1.99 ± 1.79	1.58 ± 0.65	1.558	0.122a
LDL cholesterol, mmol/L	3.02 ± 0.83	3.17 ± 0.77	0.983	0.328a
HDL cholesterol, mmol/L	1.08 ± 0.29	1.18 ± 0.28	1.97	0.051a

n, number of participants; BMI, body mass index; SBP-DBP, systolic and diastolic blood pressure; LDL-HDL, low and high density lipoprotein; AST-ALT, aspartate and alanine transaminase; SCr, serum creatinine.

^aStatistical analysis by t-test.

^bStatistical analysis by chi square test.

Effects of FBT on serum lipids

At baseline, serum lipid levels, including TC, TG, LDL-C, and HDL-C, were comparable between the FBT and control groups. Within the FBT group, LDL-C significantly decreased after the 12-week intervention (mean Δ = −0.36 mmol/L, 95% CI: −0.47 to −0.26, p < 0.001), whereas no significant change was observed in the control group (mean Δ = +0.03 mmol/L, p = 0.705) (Figure 2A). HDL-C increased significantly in both the FBT group (mean Δ = +0.14 mmol/L, p < 0.001) and the control group (mean Δ = +0.08 mmol/L, p = 0.004). TG showed a decreasing trend in the FBT group (mean Δ = −0.43 mmol/L, p = 0.083) but no significant change in the control group (mean Δ = −0.05 mmol/L, p = 0.544). TC remained stable in both groups (FBT: mean Δ = +0.08 mmol/L, p = 0.270; control: mean Δ = −0.03 mmol/L, p = 0.779). To assess between-group differences, linear mixed-effects models were employed. The group × time interaction was significant for LDL-C (χ² = 16.73, p < 0.001), indicating that FBT consumption produced a significantly greater reduction over time compared with the control group (Δ = −0.40 mmol/L), independent of age, sex, alcohol intake, and smoking. For TG, the group × time interaction was not significant (χ² = 2.13, p = 0.145; Δ = −0.38 mmol/L); notably, age (p = 0.021) and sex (p = 0.032) were significant covariates, suggesting that the within-group decreasing trend in TG may be partially confounded by these factors. The group × time interactions were also not significant for TC (χ² = 0.82, p = 0.364; Δ = +0.10 mmol/L) or HDL-C (χ² = 3.77, p = 0.052; Δ = +0.06 mmol/L). None of the covariates were significant predictors for LDL-C (Supplementary Table 3).

Figure 2.

Effects of Fuzhuan brick tea on serum lipids (A) and vascular function (B) in healthy adults. (A) Serum levels of total cholesterol, triglycerides, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol before and after the 12-week intervention in the FBT and control groups. (B) Flow-mediated dilation and nitroglycerin-mediated dilation before and after the 12-week intervention in the FBT and control groups. Data are presented as individual values with lines connecting paired samples. Within-group comparisons (before vs. after intervention) were performed using paired t-tests. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 vs. baseline within each group.

Effects of FBT on vascular function

FMD and NMD were measured to evaluate the effects of FBT on vascular function. At baseline, FMD values were lower in the FBT group than in the control group (10.97 ± 5.53 vs. 14.39 ± 8.52%). Within the FBT group, FMD significantly increased after the 12-week intervention (mean Δ = +5.45%, 95% CI: 3.14–7.76, p < 0.001), whereas no significant change was observed in the control group (mean Δ = −1.21%, 95% CI: −2.89 to 0.46, p = 0.153) (Figure 2B). Similarly, NMD significantly increased in the FBT group (mean Δ = +5.35%, 95% CI: 2.60–8.10, p < 0.001) but not in the control group (mean Δ = −0.71%, 95% CI: −2.51 to 1.09, p = 0.434). To assess between-group differences, linear mixed-effects models were employed. The group × time interaction was significant for both FMD (χ² = 21.89, p < 0.001; Δ = +6.66%) and NMD (χ² = 13.67, p < 0.001; Δ = +6.06%), indicating that FBT consumption produced significantly greater improvements in vascular function over time compared with the control group, independent of age, sex, alcohol intake, and smoking. None of the covariates were significant predictors for either FMD or NMD (Supplementary Table 3). To address the baseline difference in FMD between groups, we performed ANCOVA with week-12 FMD as the dependent variable and baseline FMD as the covariate. After adjusting for baseline FMD, the between-group difference remained significant (adjusted mean difference: 6.27%; 95% CI: 3.40, 9.14; p < 0.001; Cohen’s d = 0.59). Further adjustment for age, sex, and BMI yielded consistent results (adjusted mean difference: 6.45%; 95% CI: 3.44, 9.45; p < 0.001; Supplementary Table 4). Similarly, ANCOVA for NMD showed a significant between-group difference (adjusted mean difference: 5.92%; 95% CI: 2.63, 9.20; p < 0.001). These findings confirm that the observed vascular function improvements are not attributable to baseline imbalance.

Effects of FBT on gut microbiota richness, diversity, and composition

Shotgun metagenomics sequencing was conducted to determine the impact of FBT on gut microbiota composition. On average, 14.81 ± 4.09 Gb of raw reads were obtained per sample. Across all samples, we identified 16 phyla, 34 classes, 60 orders, 118 families, 373 genera, and 1,019 species. Alpha diversity, measured using Shannon, inverse Simpson, and Simpson indices, showed no significant differences between groups at any taxonomic level throughout the study (Figure 3A). This finding was confirmed by longitudinal analysis using SplinectomeR, indicating that FBT did not affect alpha diversity (Figure 3A).

Figure 3.

Impact of Fuzhuan brick tea on the gut microbiome over time. (A) Longitudinal trajectories of Shannon index, InvSimpson index, and Simpson index, indicating alpha (within-sample) diversity over time; p values are from trend comparison using SplinectomeR (999 permutations). (B) PCoA depicting beta (between-sample) diversity based on Bray–Curtis dissimilarity at each time point, with confidence ellipses for each group; p values are from PERMANOVA (99,999 permutations) adjusted for age, gender, alcohol intake, and smoking. (C) Longitudinal trajectories of the relative abundances of 6 SCFA-producing bacteria over time; p values are from trend comparison using SplinectomeR (999 permutations). The lines represent the group splines for the FBT group (red) (n = 55) and the control group (blue) (n = 55). Species-level longitudinal differences were assessed using permuspliner (unadjusted p < 0.05); these findings are exploratory and did not survive FDR correction across all tested species (see Supplementary Table 5).

Principal coordinates analysis (PCoA) of gut species profiles revealed subtle but statistically detectable shifts in bacterial community composition over the intervention period (Figure 3B). At baseline, no significant differences in beta diversity (community composition) were observed between groups after adjustment for age, gender, alcohol intake, and smoking (R² = 0.011, p = 0.134). Small but significant group differences emerged at Week 6 (R² = 0.016, p = 0.028) and Week 12 (R² = 0.019, p = 0.01), indicating that FBT consumption was associated with modest alterations in overall community composition. The betadisper test confirmed no significant difference in multivariate dispersion between groups at week 12 (F = 0.9931, p = 0.321), supporting the reliability of the PERMANOVA findings. The predominant phyla included Firmicutes, Bacteroidetes, Actinobacteria, and Proteobacteria. Although baseline composition was similar between groups, the relative abundance of Actinobacteria was lower, and Verrucomicrobia was higher in the FBT group than in the control group (Supplementary Figure 1).

We also examined longitudinal species-level differences and identified 77 bacterial species with nominally significant differences between groups (unadjusted p < 0.05; Supplementary Figure 2). After Benjamini-Hochberg FDR correction across all tested species, none remained statistically significant (Supplementary Table 5). Therefore, all species-level findings reported herein should be interpreted as exploratory. Among these exploratory findings, the FBT group exhibited a discernible upward trend in the relative abundance of several short-chain fatty acid (SCFA)-producing bacteria, including Roseburia hominis, Eubacterium saphenum, Eubacterium sp. AF15 50, Alistipes dispar, and Blautia stercoris (Figure 3C). Additionally, Lactobacillus amylolyticus was more abundant in the FBT group, showing a suggestive longitudinal difference compared to the control group (Figure 3C).

Effects of FBT on the functional characterization of gut microbiota

We also investigated potential functional changes in gut microbiota. To estimate the abundance of metabolic pathways, we mapped our metagenomic reads to functional orthologues from the KEGG and GMM databases. Using Bray–Curtis dissimilarity-based PCoA, we analyzed variations in 103 GMM metabolic pathways across different groups and time points. This analysis revealed a subtle but statistically detectable difference in pathway distribution between the two groups at Week 12 (Figure 4A).

Figure 4.

Impact of Fuzhuan brick tea on the functional characterization of the gut microbiome in healthy adults over time. (A) PCoA depicting the differences in functional characterization based on Bray–Curtis dissimilarity at each time point; P values are from PERMANOVA (99,999 permutations) adjusted for age, gender, alcohol intake, and smoking. (B) Bar chart depicting metabolic pathways showing nominal differential abundance (permuspliner p < 0.05) between the FBT and control groups over the 12-week intervention; Pathways are ordered by permuspliner p-value.

We further employed SplinectomeR for a longitudinal analysis of GMM pathway abundance between the FBT and control groups. Fourteen metabolic pathways showed nominal differences in abundance (permuspliner p < 0.05) (Figure 4B). Specifically, the FBT group showed enrichment in seven pathways: cysteine biosynthesis/homocysteine degradation, glycine degradation, the Entner-Doudoroff pathway, pyruvate lyase, phenylalanine degradation, propionate production III, and threonine degradation I. Conversely, the control group showed higher abundances in another set of seven pathways, including pectin degradation II, lactate production, glyoxylate bypass, succinate production, galacturonate degradation I, and ethanol production I. However, none remained significant after Benjamini–Hochberg false discovery rate (FDR) correction (Supplementary Table 6). These pathway-level findings should therefore be interpreted as exploratory.

The gut metagenome is associated with lipid profile and vascular function

Lastly, we explored the correlations between gut microbial taxa, functions, and clinical parameters, including lipid profile and vascular function. We performed a three-pronged correlation analysis between microbial species, functions, and clinical parameters using Spearman’s rank correlation (Figure 5). Among the 77 differentially abundant bacterial species (exploratory, unadjusted p < 0.05), 22 species showed significant correlations with lipid profile parameters, while 10 species were significantly associated with vascular function. Five SCFA-producing bacteria, including Roseburia hominis, Eubacterium saphenum, Eubacterium sp. AF15 50, Alistipes dispar, and Blautia stercoris, exhibited significant negative correlations with LDL-C levels, which were enriched in the FBT group. Eubacterium saphenum and Lactobacillus amylolyticus were not only negatively correlated with LDL-C levels but also positively correlated with FMD and NMD. Furthermore, the propionate production III pathway (one of the nominally differentially abundant GMMs) showed a negative correlation with LDL-C levels and a positive correlation with FMD and NMD. A sensitivity analysis restricted to week-12 data alone produced a correlation pattern highly consistent with the primary analysis (Supplementary Figure 3).

Figure 5.

Three-pronged association heatmap of bacterial species, GMM pathways, and clinical parameters. The left panel denotes the family-level and phylum-level taxonomy and directionality of enrichment in the FBT group (red) and control group (green) for species. The centered heatmap denotes the Spearman correlation between each species and each GMM pathway (red: positive correlation; blue: negative correlation). The right panel denotes the Spearman correlations between bacterial species and clinical parameters, colored by positive (pink), negative (green), or nonsignificant correlation (white). The bottom panel denotes the significant association between the GMM pathways and clinical parameters (cyan: positive correlation; brown: negative correlation). ‘*’ denotes FDR p < 0.05; ‘**’ denotes FDR p < 0.01.

Discussion

In the current study, we observed that daily FBT consumption for 12 weeks exhibited significant potential in modulating serum lipid levels, notably reducing LDL-C levels. Furthermore, FBT consumption enhanced both endothelium-dependent and endothelium-independent vascular dilation functions, indicating a beneficial impact on vascular health. Additionally, our results revealed that FBT consumption led to favorable shifts in the gut microbiota composition, with nominally enriched SCFA-producing taxa and related functional modules; although these taxonomic and pathway-level differences did not survive FDR correction, the directionality of the findings is consistent with a potential contribution of SCFA-related microbial metabolism to the observed cardiometabolic benefits.

Extensive evidence from multiple RCTs have consistently shown that tea consumption can significantly improve blood lipid metabolism. A comprehensive meta-analysis, including 10 RCTs with 411 individuals, found that black tea consumption led to a significant reduction in LDL-C levels by an average of 4.64 mg/dL, without markedly affecting TC and HDL-C levels (31). Complementing these findings, another meta-analysis of 31 trials involving 3,321 subjects reported that green tea consumption similarly resulted in a significant decrease in both TC and LDL-C levels (32). However, green tea also reduced TG levels, unlike black tea, did not reduce HDL-C levels. In addition, this analysis did not display any significant publication bias. In a randomized, double-blind study by Venkatakrishnan et al., catechin-enriched oolong tea significantly reduced TC levels by 8.89%, LDL-C levels by 6.30%, and TG levels by 10.4%, while increasing HDL-C levels by 1.06-fold in mildly hypercholesterolemic subjects over 12 weeks, outperforming green tea and the placebo in lipid regulation (33). Although a plethora of animal-based research has indicated that FBT markedly ameliorates lipid metabolism, mainly through a reduction in TG and LDL-C levels, RCTs supporting these findings in humans are limited. The only previously published work by Fu et al., encompassing a 120-day regimen of FBT on a small cohort of 10 individuals, demonstrated a decrease in TG, LDL-C, and TC levels, concomitantly with an elevation in HDL-C levels. Our current study partially aligns with these findings, demonstrating a robust reduction in LDL-C that was confirmed by linear mixed-effects modeling independent of confounders (34). This suggests that the lipid-lowering effects of FBT may be attributable to its unique composition, including polyphenols, polysaccharides and Eurotium cristatum, which might modulate lipid metabolism through various mechanisms.

Tea consumption has also been linked to improved vascular function and a reduced risk of cardiovascular diseases (5). A meta-analysis by Ras et al. (35) highlighted that tea intake boosted FMD, a key marker of vascular health, by 2.6%. Grassi et al. (36) further demonstrated that black tea significantly enhanced FMD and reduced arterial stiffness in healthy individuals. Additionally, a study by Alexopoulos et al. (37) involving 14 participants showed that green tea acutely improved endothelial function, as evidenced by increased FMD but not NMD. Our study contributed to this evidence by showing that a 12-week intervention with FBT not only improved endothelium-dependent FMD but also enhanced non-endothelium-dependent vasodilation (NMD) in healthy adults. The improvements in vascular function observed in our study may be attributed to FBT’s unique composition. These results suggest that FBT, like other tea types, offers cardioprotective benefits, possibly through mechanisms involving endothelial function modulation.

The gut microbiome has been considered a potential mediator of favorable cardiometabolic effects from tea consumption (38). Compounds from tea are not easily absorbed in the small intestine, which facilitates their interaction with gut microbiota. It is generally agreed that the small intestine only absorbs a very small percentage of tea polyphenols and polysaccharides that have been ingested, typically less than 5% (39). Many tea polyphenols and polysaccharides escape absorption in the small intestine and remain intact when they reach the colon, where they are exposed to more than 1,000 different kinds of gut bacteria. Gut microbes not only metabolize and break down tea polyphenols into smaller metabolites but also are also affected by tea polyphenols and their biological derivatives. Empirical evidence has shown that all six types of tea, including green tea, white tea, yellow tea, oolong tea, black tea, and dark tea, significantly modulated the gut microbiota of rats with a high-fat diet-induced metabolic syndrome, despite that specific detailed bacterial taxa may vary depending on the type of tea (40). Furthermore, these alterations exhibit a conspicuous correlation with the amelioration of metabolic syndrome. In vitro incubation of human gut microbiota has revealed that tea polyphenols and polysaccharides can directly and selectively modulate the abundance of specific bacterial species, significantly increasing Bifidobacterium and Lactobacillus while inhibiting the proliferation of Prevotella and Clostridium (41). Proof-of-principle studies from animal experiments using antibiotics and fecal transplants support the concept that gut microbiota acts as an antecedent for tea extracts to improve cardiometabolic health. In a study by Lu et al., the transplantation of fecal microbiota from donor mice treated with an extract of ripened Pu-erh tea exhibited a notable impact on the recipient mice’s modulation of weight and metabolic syndrome symptoms (42). Consistently, Tang et al. (43) reported that the extract of Liupao tea prevented metabolic-associated fatty liver disease induced by a high-fat diet in ApoE^−/− mice through fecal microbiota transplantation. Noteworthy, Liu et al. (44) provided evidence indicating that the beneficial effects of FBT in reducing obesity and its associated metabolic disorders are counteracted by the depletion of microbiota following antibiotic treatment. Further, accumulating literature has demonstrated that the unique components found in FBT, in addition to tea polyphenols, also exert significant regulatory effects on gut microbiota. For instance, Eurotium cristatum, a dominant fungus in FBT production, significantly alters gut fungal and bacterial composition and increases butyrate levels fivefold in mice, thereby reducing obesity and underscoring its probiotic benefits for gut health (45, 46). The purified fraction of polysaccharides derived from FBT, specifically FBTPS-2 and FBTPS-3, also exerted anti-inflammatory and anti-obesity effects by modulating gut microbiota and enhancing SCFA production (47–50). The above evidence was mainly derived from animal experiments, while research in humans has been limited. Yuan et al.’s study on healthy individuals revealed that green tea liquid consumption can modify both oral and gut microbiota compositions (23). However, a 12-week RCT failed to observe any significant effects of green tea polyphenols on the gut microbiota of healthy participants (22). Similarly, in a randomized, placebo-controlled, double-blinded trial involving healthy postmenopausal women, fecal microbiota was not significantly influenced following a 12-month administration of a green tea polyphenol supplement (21). The inconsistency may be attributed to factors such as the study design, sample size, dosage, compliance, and depth of 16S rRNA sequencing. To the best of our knowledge, this is the first RCT to investigate the effects of FBT on gut microbiota. In this study, we observed subtle but statistically detectable shifts in the composition of the bacterial community in healthy individuals following FBT consumption, which is in line with a previous work by Huang et al., highlighting the modulatory effect of post-fermented Pu-erh tea on gut microbiota in 13 healthy male volunteers (51). Our study extends the concept that FBT may affect gut microbiota and cardiometabolic health in humans.

In the current study, exploratory species-level analyses suggested an increased abundance of several bacteria with known beneficial properties, including Blautia stercoris, Roseburia hominis, Eubacterium saphenum, Eubacterium sp. AF15 50, Alistipes dispar, and Lactobacillus amylolyticus, all of which can produce SCFA from carbohydrate metabolism. Although these species-level findings did not survive FDR correction, the convergent direction of change across multiple SCFA-producing taxa and their consistent correlations with clinical outcomes (after FDR correction) lend biological plausibility to these exploratory observations. Consistently, these observations were supported by the increase in potential functional modules associated with propionate production. Blautia, a genus of the Lachnospiraceae family, plays an important role in maintaining intestine homeostasis, preventing inflammation, and promoting the metabolism of unbound lipids and glucose by producing SCFAs. The abundance of Blautia is negatively correlated with metabolic diseases, such as obesity, nonalcoholic fatty liver, and type 2 diabetes (52). Roseburia hominis, a butyrate-producing species that restores intestinal barrier function and regulates immune function, is found to be decreased in ulcerative colitis, hypertension, alcoholic fatty liver, and type 1 diabetes (53). Eubacterium is another SCFA-producing bacterium inversely associated with liver total TG, interleukin-2, and C-reactive protein levels. Alistipes is a propionate producer that plays a protective role in cardiovascular diseases, such as hypertension, atrial fibrillation, heart failure, and atherosclerosis cardiovascular disease (54). Lactobacillus, a famous probiotic, is known to produce bile salt hydrolase and SCFAs to inhibit cholesterol absorption (55). Short-term Lactobacillus supplementation has been shown to improve vascular endothelial function and reduce systemic inflammation (56). Previous animal experiments have also corroborated that administration of crude FBT extract or its bioactive constituents increases the abundance of SCFA-producing bacteria, which was accompanied by augmented SCFA levels in fecal and blood samples (48, 57–60). SCFAs play a critical role in ameliorating various risk factors associated with CMDs, including dyslipidemia, insulin resistance, inflammation, endothelial dysfunction, and vascular remodeling (61). This beneficial impact may be attributable to their ability to inhibit histone deacetylases and activate diverse free fatty acid receptors (62). We further explored potential interactions between gut microbiota and clinical outcomes. Our findings revealed significant negative correlations between these SCFA-producing bacteria and LDL-C levels, while positive correlations were observed with FMD and NMD. Taken together with previous studies, our data suggest that an increase in these SCFA-producing bacteria may partially contribute to the beneficial effect of FBT consumption on cardiometabolic health, although these species-level observations require confirmation in larger, independent cohorts.

The primary strength of our study lies in the RCT design, which established a robust framework for systematic data collection. Several limitations should be acknowledged. First, due to the distinct aroma and taste of FBT, it is not feasible to implement a blinded intervention among participants, which could potentially introduce expectation bias. A larger, double-blind, placebo-controlled randomized trial is needed to validate our findings. Second, while we employed metagenomic sequencing to detect species-level differences, our study did not include fecal or serum SCFA quantification, as the sample collection protocol was not optimized for metabolomic analysis of volatile short-chain fatty acids. Consequently, while our metagenomic data and functional pathway analysis support an enrichment of SCFA-producing bacteria and related biosynthetic pathways, we cannot directly confirm increased SCFA production or establish a causal link between SCFA-producing bacteria and the observed cardiometabolic improvements. Future studies incorporating targeted SCFA metabolomics with appropriate sample preservation protocols are warranted to validate this proposed hypothesis. Third, this study does not allow us to conclude the direct causality between gut microbiota alterations and the observed health benefits, which should be further verified by germ-free animals and fecal transplant experiments. Fourth, the 12-week intervention may be insufficient for the gut microbiota to reach a new steady state, and the lack of post-withdrawal follow-up limits our understanding of the persistence of microbiota changes after FBT cessation. Fifth, the identification of differentially abundant bacterial species and GMM pathway was based on permuspliner p-values without cross-species multiple testing correction. Although SplinectomeR’s permutation-based framework provides statistical control for longitudinal comparisons, neither the species-level nor the pathway-level findings survived Benjamini–Hochberg FDR correction, and these results should therefore be interpreted as exploratory and require validation in independent cohorts with adequate sample sizes. Finally, the composition of FBT is complex, and our study cannot decipher the involved bioactive compounds responsible for modulating gut microbiota. More work is needed to elucidate the biological activity of FBT and identify the bioactive compounds and their corresponding metabolites.

Conclusion

In conclusion, our study demonstrated a beneficial effect of FBT consumption on the lipid profile and vascular function. Exploratory species-level analyses suggested that FBT consumption is associated with favorable alterations in gut microbiota, particularly by augmenting the abundance of SCFA-producing bacteria, although these findings require validation in larger studies. This suggests a potential functional role for these microbial changes in the beneficial effects of FBT consumption in humans. These findings align with the increasing recognition of tea consumption as a low-risk, cost-effective preventive strategy against CMDs. Nevertheless, future well-designed trials with a larger sample size; a randomized, double-blind, placebo-controlled, multicenter design, alongside comprehensive multiomics analyses are needed to establish a clearer mechanistic understanding of the potential crosstalk between FBT consumption and gut microbiota, and their impacts on host cardiometabolic health.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by the grants from the Key Industrial Innovation Chain Project in Shaanxi Province of China (grant no. 2021ZDLSF02-03), Key Program for the Traditional Chinese Medicine Inheritance and Innovation and ‘Qin Medicine’ Development (grant no. 2021-03-22-001), the Key Basic Natural Science Foundation of Shaanxi Province (grant no. 2022JZ-47), Qin Chuangyuan Traditional Chinese Medicine Innovation Research and Development Transformation Project (grant no. 2022-QCYZH-022).

Data availability statement

The metagenomic sequencing data presented in this study have been deposited in the NCBI BioProject Database under accession code PRJNA1083321. (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1083321).

Ethics statement

The studies involving humans were approved by Ethics Committee of Shaanxi Provincial People’s Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent was obtained from all participants.

Author contributions

FL: Investigation, Formal analysis, Writing – original draft, Funding acquisition, Conceptualization. YLv: Conceptualization, Writing – original draft, Investigation, Project administration. TW: Data curation, Methodology, Writing – original draft. CL: Formal analysis, Writing – review & editing. YZ: Writing – review & editing, Formal analysis. LL: Data curation, Writing – review & editing. XH: Supervision, Conceptualization, Writing – review & editing. YLi: Writing – review & editing, Conceptualization, Supervision. FZ: Writing – review & editing, Project administration, Supervision. JW: Supervision, Writing – review & editing, Funding acquisition, Conceptualization, Project administration, Data curation.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. During the preparation of this work the authors used ChatGPT for English language optimization. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2026.1827773/full#supplementary-material

SUPPLEMENTARY FIGURE 1

Impact of Fuzhuan brick tea on the gut microbiome composition at the phylum levels in healthy adults over time.

SUPPLEMENTARY FIGURE 2

Longitudinal differences in the bacteria between the two groups over time using the SplinectomeR permuspliner function. These species were identified using permuspliner with unadjusted p < 0.05. After Benjamini-Hochberg FDR correction across all 77 species, none remained statistically significant (see Supplementary Table 5). All species-level findings should be interpreted as exploratory.

SUPPLEMENTARY FIGURE 3

Three-pronged association heatmap of bacterial species, GMM pathways, and clinical outcomes at week 12 alone. The left panel denotes the family-level and phylum-level taxonomy and directionality of enrichment in the FBT group (red) and control group (green) for species. The centered heatmap denotes the Spearman correlation between each species and each GMM pathway (red: positive correlation; blue: negative correlation). The right panel denotes the Spearman correlations between bacterial species and clinical parameters, colored by positive (pink), negative (green), or nonsignificant correlation (white). The bottom panel denotes the significant association between the GMM pathways and clinical parameters (cyan: positive correlation; brown: negative correlation). * denotes FDR p < 0.05; ** denotes FDR p < 0.01.