Participants (N = 240) aged over 60 years were invited to take part by donating a fecal sample between 2017 and 2022 at the Hospital Centro Medico Nacional Siglo XXI (CMN-IMSS), All participants signed an informed consent form. The inclusion criteria consisted of male and female individuals aged 60 years and older, diagnosed with hypertension, and receiving antihypertensive treatment for at least 4 years at the time of recruitment. The exclusion criteria included diagnosis of cancer, chronic neurodegenerative or immune diseases, and insufficient or deficient sample quality. The protocol was approved by the Committees of Research Ethics under approval numbers R2018-785-004 and CEI2017/04 & 23/2016/I. This research followed current bioethical and safety regulations, including the principles of the Declaration of Helsinki. Fecal samples were collected from patients who were carefully instructed and provided with an in-house collection kit. Samples were added in RNA-later (Thermo-Scientific) and stored at −70°C until DNA extraction. Blood pressure was measured three times within 15–30 min using a sphygmomanometer. The patients were then classified as controlled (<140/90 mmHg) and uncontrolled (≥140/90 mmHg) according to institutional guidelines. This study focused on comparing these groups based on all available clinical and demographic data; however, it did not include a normotensive group or its lifestyle habits.

DNA was isolated from 200 mg of feces using the QIAamp Fast DNA Stool Mini Kit (Qiagen, United States) according to the manufacturer’s instructions and stored at −20°C. The hypervariable region V3-V4 was amplified using the 16S V3 (341F) forward and V4 (805R) reverse primers and adapters from Illumina following the manufacturer’s 16S metagenomic sequencing library protocol. PCR reactions were 30 μL in volume, containing 4 μL of the DNA (50 ng/μL), 0.25 μL of each PCR primer (10 pM), and 15 μL of 2X Platinum™ SuperFi™ PCR Master Mix (Invitrogen, United States). Amplification was performed for 25 cycles consisting of 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s, followed by a final extension at 72°C for 5 min. The fragments were cleaned with Agencourt AMPure XP beads (Beckman Coulter Genomics, Brea, CA, United States) according to the manufacturer’s protocol. Indexes and adaptors were ligated by PCR with 5 μL of Illumina Nextera XT Index Primer 1 (N7XX), 5 μL of Nextera XT Index Primer 2 (S5XX), and 25 μL of 2X Platinum™ SuperFi™ PCR Master Mix (Invitrogen, USA) in a thermocycler at 95°C for 3 min, as well as six cycles at 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s, and a final extension at 72°C for 5 min. The 16S rRNA V3-V4 libraries were purified with Agencourt AMPure XP beads. Library quality control was verified by microcapillary electrophoresis using a TapeStation 4,200 (Agilent Technologies, CA, United States). Then, the libraries were normalized and pooled to 10.2 nM, denatured, and diluted to a final concentration of 10 pM, including 20% of PhiX. The libraries were sequenced using a 2x250bp cartridge/MiSeq Reagent Kit V3 in a MiSeq sequencer (Illumina).

Sequencing paired-end FASTQ files were evaluated for quality control using QIIME2 v2024.5, followed by denoising with the Divisive Amplicon Denoising Algorithm 2 (DADA2) plugin. The resulting amplicon sequence variants (ASVs) were used to generate a taxonomy table with a naive Bayes pre-trained classifier for the V3-V4 hypervariable region of the 16S rRNA gene, based on the ribosomal database SILVA_138. QIIME2 artifacts were imported into R using the qiime2R package and analyzed with the Phyloseq package (Hall and Beiko, 2018). Statistical analyses were conducted in R version 4.0.4 (R Core Team, 2015).

To assess batch effects and batch correction, we conducted principal coordinate analysis (PCA) on adjusted and unadjusted rarefied relative abundance data with centered log-ratio normalization using the microViz R library. Rarefaction was set to the minimum sampling depth across samples, which was 26,500 sequences per sample. The core microbiome was assessed based on a sample prevalence of > 50% at a relative abundance frequency of > 1% at the genus level. Several alpha diversity indexes were assessed, including the observed species, Shannon index, Chao index, Simpson dissimilarity, and Fisher index. Significant differences in alpha diversity across the groups were calculated using the Kruskal–Wallis and Wilcoxon tests. PCA and redundancy analysis (RDA) were conducted with centered log-ratio normalized counts at the genus level using the microViz R library. The RDA included clinical variables, such as age, uncontrolled systolic or diastolic blood pressure (DBP), cholesterol, HDLC, LDLC, glucose, and triglycerides. Beta diversity was calculated using Bray–Curtis dissimilarity distances. In addition, a permutational multivariate analysis of variance (PERMANOVA) was conducted on 999 permutations to test the association between the composition of the microbiota and clinical variables, such as SBP, DBP, age, and sex. Differential abundance analyses at the genus level were performed using the linear regression framework for differential abundance analysis (LinDA), fitting a linear model for abundance data and correcting for compositional effects and biases. p-values were adjusted using the false discovery rate (FDR) method, with a significance threshold set at a p-value of ≤ 0.01.

In addition, we investigated hypertension-linked bacterial taxa by performing linear regression analyses of bacterial abundance and blood pressure control. Bacteria abundances were CLR-transformed to account for compositionality, and the models were adjusted for age, sex, and diabetes status. Significance was determined at a p-value of < 0.05, with effect sizes reported as β-coefficients, reflecting blood pressure changes in mmHg per unit increase in CLR-transformed abundance. These analyses specifically tested whether previously reported hypertension-associated taxa showed consistent relationships with blood pressure gradients in our cohort, irrespective of clinical control status.

Finally, functional prediction and differential abundance analysis were performed to predict functional profiles from ASVs using PICRUSt2 (v2.5.0). Predictions were reported as enzyme commission (EC) numbers and KEGG orthologs (KOs). Functional predictions were analyzed using DESeq2 (v1.40.0), comparing uncontrolled (SBP ≥ 140 mmHg or DBP ≥ 90 mmHg) versus controlled (SBP < 140 mmHg or DBP < 90 mmHg) groups. Features with a log2FC of > 1 and an FDR-adjusted p-value of < 0.05 were considered significant. Data were normalized using a variance-stabilizing transformation (VST). Confounding factors, such as age, sex, and T2D status, were included as covariates in the DESeq2 model. Analyses were performed separately for EC numbers and KOs to identify hypertension-associated metabolic shifts.

First, we investigated the relative abundance of bacterial phyla and genera, observing that Bacteroidetes showed the highest abundance (49%), followed by Firmicutes (42%), Proteobacteria (7%), and Verrucomicrobia (1.0%). At the genus level, Bacteroides (27%), Prevotella 9 (14%), Faecalibacterium (5.4%), Lachnospiraceae, Escherichia-Shigella (4.7%), Allistipes (4.2%), Ruminococcaceae UCG-002 (3.7%), Parabacteroides (2.9%), Eubacterium coprostanoligenes (2.5%), Roseburia (1.8%), and Christensenellaceae R-7 (1.1%) were the most abundant (Figure 1). No significant differences in microbiota composition according to hypertension control for systolic and diastolic blood pressure (SBP and DBP) were observed (Supplementary Figure 1).

One of the aims of the study was to characterize the core bacterial composition of the gut microbiome in hypertensive older adults. The prevailing bacteria, ranked by prevalence and abundance, were Bacteroides, Prevotella 9, Faecalibacterium, Alistipes, Ruminococcaceae UCG-002, Parabacteroides, Eubacterium coprostanoligenes, and Roseburia, showing a prevalence of up to 40% (Figure 2). In addition, bacteria with lower but consistent abundance across the study population included Escherichia-Shigella, Paraprevotella, Phaseolarctobacterium, Ruminococcus 2, Subdoligranulum, Dialister, Ruminococcaceae UCG-014 & UCG-005, Ruminococcus 1, Christensenellaceae R-7, Barnesiella, and Blautia, each with an abundance around 20%.

We evaluated bacterial richness and diversity using several metrics, considering age as a continuous variable, and stratifying by age group. Alpha diversity indexes were compared between the controlled and uncontrolled patients. For the age group 60–75y, the indexes—Chao 1, Fisher, and observed species OTUs—showed higher richness, including more singletons and rare bacteria, compared to the individuals older than 75 years. After 80 years of age, there was an apparent decrease in bacterial richness (Figure 3). Alpha diversity indexes comparing the controlled and uncontrolled SBP and DBP groups did not show significant differences, and these groups seemed similar in terms of abundance and richness. However, there was a lower number of ASVs in the uncontrolled DBP group, as shown by a lower Chao1 index (Supplementary Figure 2).

To assess differences in bacterial composition between controlled and uncontrolled hypertension, we estimated a Bray–Curtis dissimilarity index but found no clustering differences when comparing blood pressure levels ≥140/90 mmHg with <140/90 mmHg (Supplementary Figure 3). Hence, we evaluated the impact of clinical variables using RDA, which may explain variation in hypertension control. Ruminococcaceae and Muribaculaceae seemed to partly accompany uncontrolled hypertension, more apparently for diastolic than for systolic blood pressure (Figure 4). Age was associated with changes in the abundance of Escherichia-Shigella, while variations in blood glucose and lipid levels corresponded with changes in Prevotella 9 and Phascolarctobacterium abundance. The correlation analyses between hypertension control and bacteria, including Escherichia-Shigella, did not show a relevant relationship despite previous reports (y ≤ 0.02 ~ 0.00002*BP(x), R2 ≤ 0.002, p ≥ 0.50).

To identify bacteria that could explain the differences between controlled and uncontrolled hypertensive patients, we performed a linear decomposition analysis (LinDA) and examined associations between the gut microbiome and hypertension control for SBP and DBP separately (Zhou et al., 2022). We found that when hypertension is controlled, there is an increased abundance of Ruminococcaceae UCG002 and DTU 089 and a decreased abundance of Dorea, Lachnospiraceae UCG-010, Eubacterium hallii, and Prevotella 7 (Figure 5). Abundance differences were similar in SBP and DBP for Ruminococcaceae UCG002 and DTU 089, but after correction using an FDR test, statistical significance was observed only for DBP.

Functional analyses of metagenomic predictions based on bacterial abundance differences between controlled and uncontrolled hypertension may provide information on macromolecules and their routes, potentially further explaining the relationship between the gut microbiome and hypertension control. We observed three major gene-enzyme differential changes: (a) a predicted increase in genes and enzymes related to the production of reactive oxygen species in the uncontrolled patients, including, EC1.6.3.3, K17870, and K180022; (b) a decrease in predicted genes and enzymes related to nucleic acid metabolism, including, EC2.1.1.215., K18846, K00555, and K19174/75; and (c) an increase in predicted genes and enzymes related to the biotransformation of endogenous and xenobiotic compounds, including conjugation, transferases, and multidrug-resistant enzymes and genes such as EC2.4.1.19, EC2.1.1.180, EC2.1.1.215, K18908, K15546, and K00701 (Supplementary Table 1).

The composition of the gut microbiome in older adults is highly variable, yet certain patterns have been consistently reported. For instance, Bacteroidetes and Firmicutes dominate 80% of the gut microbiome, with an increase in Proteobacteria and Escherichia-Shigella with age (Novelle et al., 2025; Salazar et al., 2017). Our findings align with previous observations, as we observed that Firmicutes and Bacteroidetes together comprise up to 91% of the gut microbiota, with Bacteroidetes being more prevalent (49%) than Firmicutes (41%; Bradley and Haran, 2024).

The core bacteria belonged to the Bacteroidetes phylum, including Bacteroides, Parabacteroides, Prevotella 9, and Alistipes, as well as to the Firmicutes phylum, including Faecalibacterium, Ruminococcaceae UCG-002, Eubacterium coprostanoligenes, and Roseburia, together composing up to 40% of the gut microbiome (Figure 2). This is consistent with previous reports on the gut microbiome of older adults (O’Toole and Claesson, 2010).

Several of the above-mentioned bacteria have been associated with aging. For example, E. coprostanoligenes has been detected in older adults, and its abundance increases with age and has been observed in the transition from adulthood to older age (Biagi et al., 2010; Sepp et al., 2022; Wei et al., 2021). Hence, the presence of E. coprostanoligenes was not surprising and could be related to the age of this population. The presence of Bacteroides, Parabacteroides, Alistipes, and Prevotella 9 characterizing the gut microbiome of the older adults in this study further confirms their role as major components of the gut microbiome, as reported in previous studies (Biagi et al., 2012; O’Toole and Claesson, 2010). Potential roles for each genus have been discussed before. Prevotella 9 is associated with accelerated aging and the inflammaging phenotype (Singh et al., 2024). The role of Parabacteroides in older adults is inconclusive. Some studies have found it to be more abundant in patients with Alzheimer’s disease, while others have associated it with positive outcomes related to mental health and diet. Its abundance seems to accompany that of several strains of Ruminococcaceae, which was also observed here (Molinero et al., 2023; Olson et al., 2018). Alistipes, another age-dependent bacterium, has been associated with anti-inflammatory properties, SCFA production, and beneficial gut health. Nevertheless, its overabundance has been reported in inflammatory bowel disease and hypertension, highlighting the fact that its roles are not fully clear and cannot be interpreted in isolation (Tokarek et al., 2023).

Furthermore, in characterizing the microbiome of hypertensive older adults, we observed several bacteria with an average abundance of 20%, including Ruminococcaceae, Phascolarctobacterium Subdoligranulum, Dialister, Blautia, Barnesiella, Paraprevotella, Christensenellaceae R-7, and Escherichia-Shigella, all of which have been reported to increase with age (Farsijani et al., 2024; M et al., 2022). The presence of Escherichia-Shigella is causative of the “inflammaging” phenotype, possibly mediated by its glycerophospholipid metabolism, which activates toll-like receptors, increasing inflammation and endothelial dysfunction (Wang et al., 2021). It is well documented that both Escherichia-Shigella and hypertension risk increase with age (Mushtaq et al., 2019). While in this study, we observed a higher abundance of Escherichia-Shigella with age, we did not observe a relationship between Escherichia-Shigella and increased blood pressure. This discrepancy may be due to age-related microbial dynamics. Our study population consisted predominantly of individuals aged 70-80 years, while most research documenting the increase of Escherichia-Shigella in hypertension tends to include a broader age range and younger individuals (Wan et al., 2021). It is possible that age itself (via inflammaging) may override or mask the distinctions based on hypertension control status. Escherichia-Shigella may already be elevated due to age in both groups, limiting our ability to detect differences associated with blood pressure control.

Variation in gut microbial diversity with age and hypertension control was analyzed using several metrics. Consistent with current research, we observed a decline in gut bacterial diversity with age (Renson et al., 2020), and a clearer trend was observed in individuals aged 75 y and older (Figure 3). Interestingly, the 65-70y age group displayed increased diversity, which persisted up to age 75. This suggests that the transition between 65 years and 70 years could represent a window for enhancing and protecting microbial diversity before its age-related decline. The uncontrolled DBP group showed lower diversity compared to the controlled DBP group (p < 0.05) for the Chao 1 index, supporting the notion that healthier phenotypes show a richer gut microbiota (Louca et al., 2021).

Several metrics were assessed to investigate microbial differences according to hypertension control. Average bacterial composition was not different between the controlled and uncontrolled patients, but Ruminococaceae UCG-002 and the Muribaculaceae family were able to explain variation in uncontrolled hypertension when considering clinical parameters (Figure 4). Ruminococcaceae UCG-002 has been negatively associated with heart disease (Li et al., 2022) and metabolic syndrome (Wutthi-in et al., 2020). The beneficial effects of the Ruminococcaceae family are related to the production of SCFAs, which modulate blood pressure through the kidney receptors—GPR41, GPR43, and GPR109A (Gao et al., 2021; Tilves et al., 2022). Ruminocococcus can also improve linoleic acid and glucose absorption, insulin sensitivity, and intestinal integrity in mice, which together may support cardiovascular health (Xie et al., 2022). Muribaculaceae abundance seems to accompany uncontrolled DBP (Figure 4). This family has been associated with both anti-inflammatory and pro-inflammatory effects depending on the environment (Zhu et al., 2024). Therefore, its role in uncontrolled DBP would require targeted experiments.

We found interesting differences in bacteria associated with controlled blood pressure that have not been previously acknowledged. For example, increases in Ruminococcaceae UCG002, Ruminococcus 2, and DTU 089—the latter being a member of the Lachnospiraceae family—have been associated with skeletal metabolism and low protein intake in older adults (Farsijani et al., 2022; Okoro et al., 2023). In this study, we validated the presence of DTU 089 in older adults, inferring that its potential influence on health could be related to diet. Ruminococcaceae UCG002 has been reported to negatively influence cardiovascular health and hypertension (Chen et al., 2022; Miao et al., 2024; Qin et al., 2022). However, most members of the Ruminococcaceae family ferment fiber to produce SCFAs and influence gut integrity, immunity, and cholesterol transport. Therefore, future research should focus on specific species and strains to elucidate their clear role (Sanders et al., 2019; Vojinovic et al., 2019).

On the other hand, controlled blood pressure was associated with decreased abundance of Dorea, Prevotella 7, Lachnospiraceae UCG010, and the Eubacterium hallii group (Figure 5). Except for Prevotella 7, these bacteria belong to the phylum Firmicutes and the family Lachnospiraceae, and members of this family can show contrasting roles in health and disease (Vacca et al., 2020). Dorea has been associated with hypertension and its complications (Miao et al., 2024), and Lachnospiraceae UCG10 has been linked as a causal factor in venous thromboembolism (Huang et al., 2025). Eubacterium hallii, also known as Anaerobutyricum hallii, is consistently reported as beneficial, since it can produce butyrate, reuterin, and vitamin B12 (Engels et al., 2016). It has also been suggested as a probiotic for cancer prevention, as it can conjugate carcinogenic pyrimidine derivatives from cooked meat (Engels et al., 2016). The beneficial properties of these members of the Lachnospiraceae family are consistent with their decreased abundance in a hypertensive population. Investigations should decipher metabolic pathway differences within family members so that specific Lachnospiraceae species can be pinpointed as more sensitive or beneficial for blood pressure control and to clarify if these associations could lead to hypertension development or protection pathways. Of special interest is the Enterobacter hallii group, whose abundance has been associated with age (Biagi et al., 2010), hypertension (Verhaar et al., 2020), inflammation, and insulin sensitivity (Udayappan et al., 2016). Its decrease in hypertensive older adults may reflect a direct depletion of its beneficial activity. However, the directionality of its abundance in healthy older adults, as well as the impact of medication and different geographical ancestries, remains unclear. Therefore, further research is needed in this area (Hayashi et al., 2003; Mueller et al., 2006).

Finally, functional predictions indicated that uncontrolled patients may have an increased molecular machinery for the generation of reactive oxygen species and metabolizing enzymes, along with a decrease in nucleic acid metabolism or repair. These observations are broad and do not directly reflect the gut functional environment. We can only infer that higher reactive oxygen species and poor nucleic acid maintenance are consistent with the phenotype of uncontrolled hypertensive patients. It is interesting to note that functional predictions also suggested an increase in metabolizing enzymes, some of which could be involved in phase II drug metabolism. This indicates that the gut microbiome of uncontrolled patients may be metabolizing drugs, such as hypertensives, at a higher rate, likely affecting their efficacy.

In summary, changes in the abundance of Ruminococcus UCG-002 and several members of the Lachnospiraceae family, such as Dorea, Eubacterium hallii, and Lachnospiraceae UCG-010, were significantly associated with hypertension control. These associations suggest that members of the Lachnospiraceae family may contribute to the regulation of blood pressure through fiber fermentation and SCFA production, followed by interactions with kidney receptors, renin, and lipopolysaccharides (Verhaar et al., 2020; Yan et al., 2022). In addition, these bacteria produce metabolites such as dopamine, serotonin, and norepinephrine, which favor angiotensin II release and vasoconstriction (Miao et al., 2024; Xu and Marques, 2022). It is relevant to mention that this study included only hypertensive individuals and did not include a normotensive reference, which may reflect a limitation of the study. Consequently, the directionality of the observed abundance changes may not fully align with previous studies.

Nevertheless, we depicted novel observations that, on the one hand, described the core microbiome of hypertensive older urban dwellers, confirming and expanding previous associations between blood pressure and SCFA producers such as Dorea, Ruminococcaceae UCG-01,0, and the Lachnospiraceae family. On the other hand, we conducted one of the largest studies in an underexplored admixed population, showing an association between gut bacteria and hypertension control.

These observations are among the first to describe the gut microbiome of hypertensive older adults, contributing to the growing body of evidence on the role of the microbiome and its changes in aging and hypertension. Future endeavors should include a normotensive group, as well as assessments of lifestyle and dietary habits, to broaden and validate the observed changes in the gut microbiome of hypertensive older adults.