The question of whether hypertension damages the brain has long been debated in epidemiology. Numerous observational studies have found that people with midlife high blood pressure are more likely to develop white matter lesions, cognitive decline, and dementia, including Alzheimer’s disease. A 30-year follow-up of the CARDIA study, for example, showed that elevated blood pressure in midlife predicted advanced brain aging decades later, while analyses of the UK Biobank have linked higher midlife blood pressure to more severe white matter hyperintensities—bright spots on brain scans that signal vascular damage. But observational studies cannot rule out reverse causation or confounding: perhaps early, undetected brain degeneration drives up blood pressure, or perhaps unrelated lifestyle factors influence both. To break this deadlock, the research team turned to genetics, exploiting the random assortment of genes at conception as a natural experiment.

At the heart of the study is a measure known as the brain age gap, or BAG. Neuroscientists can train machine-learning models on magnetic resonance imaging scans to predict a person’s chronological age from the structure of their brain. When the predicted “brain age” exceeds the actual age, the difference—the brain age gap—is thought to reflect accelerated aging, and a larger gap has been linked to cognitive impairment and neurodegenerative disease. The researchers drew on a genome-wide association study, or GWAS, of the brain age gap in individuals of European ancestry, compiled from the UK Biobank and the LIFE-Adult cohort and curated by Philippe Jawinski and colleagues. As the exposure side of their analysis, they meta-analyzed hypertension GWAS data from the Million Veteran Program (MVP), a vast resource covering genetic diversity across more than two thousand traits, and release 12 of the FinnGen database, which draws on a well-phenotyped Finnish population.

The first analytical step was to test whether the two traits share genetic underpinnings. The team applied linkage disequilibrium score regression (LDSC), a method that estimates genetic correlation by examining whether SNPs associated with one trait tend to also show elevated association signals with another, accounting for the confounding effect of linkage disequilibrium across the genome. They supplemented this with high-definition likelihood (HDL) inference and stratified LD score regression (S-LDSC), which partitions heritability across functional annotations of the genome. All approaches converged: hypertension and the brain age gap are significantly genetically correlated, meaning the same genetic variants that raise blood pressure risk also tend to push the brain toward an older structural profile.

Correlation, however, is not causation. To probe directionality, the researchers deployed Mendelian randomization (MR), a technique that uses genetic variants as proxies—so-called instrumental variables—for a risk factor. Because alleles are allocated randomly at conception and are fixed from birth, they are largely immune to the reverse causation and environmental confounding that plague observational research. The team used the inverse-variance weighted method alongside generalized summary-data-based Mendelian randomization (GSMR), which additionally filters out weak instruments and performs heterogeneity-in-dependent-instruments (HEIDI) tests to remove biased associations driven by linkage. The results pointed in one direction: genetic liability to hypertension exerts a causal effect that increases the brain age gap. In other words, the genes that predispose a person to high blood pressure appear to make the brain biologically older.

With causality suggested, the next challenge was to find the genes. The researchers performed transcriptome-wide association studies (TWAS), which integrate GWAS summary statistics with expression quantitative trait loci (eQTL) data from the GTEx v8 project—datasets describing how genetic variants influence gene expression in tissues. Four complementary TWAS methods were used: MAGMA, which aggregates SNP effects at the gene level; FUSION, which imputes gene expression from functional summaries; JTI-PrediXcan, a joint-tissue imputation approach; and FOCUS, which applies fine-mapping to distinguish likely causal genes from nearby correlated ones. This multipronged strategy identified 15 genes shared between hypertension and brain aging. Crucially, the team then validated 10 of them using summary-data-based Mendelian randomization (SMR), which tests whether the association between a gene’s expression and a trait is likely causal rather than a byproduct of nearby genetic linkage.

One gene rose above the rest. Using the polygenic priority score (PoPS) method, which ranks genes by integrating their functional features and patterns of polygenic enrichment, the analysis singled out TRIM47 as the core gene linking hypertension to the brain age gap. TRIM47 encodes a member of the tripartite motif family of E3 ubiquitin ligases, proteins that tag other molecules for degradation and modulate cellular signaling. Intriguingly, TRIM47 had already emerged in prior research as a strong candidate in cerebral small vessel disease, and laboratory work has shown that it acts as an endothelial activation factor, aggravating inflammatory injury through K63-linked ubiquitination of the signaling protein TRAF2. Other studies have shown that TRIM47 regulates autophagy in brain endothelial cells—the cells that form the blood-brain barrier. This convergence of genetic and functional evidence paints TRIM47 as a plausible molecular switch through which vascular stress translates into brain tissue damage.

Perhaps the most visually striking findings came from the spatial analysis. The team employed GSMAP—genetically informed spatial mapping of cells for complex traits—in combination with spatial transcriptomics, a technology that records gene expression while preserving information about where in a tissue each transcript originates. This allowed the researchers to ask not just which genes are involved, but where in the brain, and in which cell types, their activity matters. The shared hypertension–brain aging signals were enriched in the meninges, the protective membranes enveloping the brain; in fiber tracts, the brain’s white matter highways; in cortical layer 1, the outermost, input-receiving sheet of the cerebral cortex; and in the CA1 stratum lacunosum/radiatum, a hippocampal layer critical for memory processing and a region vulnerable in early neurodegeneration.

The cell-type results reinforced a vascular and myelin-centric story. Enrichment appeared in meningeal cells, vascular smooth muscle cells—which regulate vessel tone and whose stiffening is a hallmark of hypertension—oligodendrocytes, the myelin-producing cells of the central nervous system, and specific astrocyte subtypes that support the blood-brain barrier and metabolic homeostasis. Together, these enrichments sketch a mechanistic hypothesis: chronic high blood pressure stresses the cerebral vasculature and meninges, compromises endothelial and smooth muscle function, and secondarily injures myelin and the glial cells that maintain it, accelerating the structural changes that machines read as an older brain. The authors are careful to note that the spatial framework is hypothesis-generating rather than definitive, but it offers a roadmap for laboratory studies that could test each link in the chain.

The clinical implications are potentially significant. Hypertension is one of the most common and most modifiable risk factors in medicine, and if it causally accelerates brain aging—as this genetic evidence suggests—then aggressive blood pressure control, particularly in midlife, may represent one of the most effective strategies for preserving cognitive health into old age. The identification of TRIM47 and the associated meningeal, smooth muscle, oligodendrocyte, and astrocyte pathways also opens concrete therapeutic avenues: drugs that modulate TRIM47 activity, protect the blood-brain barrier, or support myelin integrity could, in principle, decouple vascular disease from neurodegeneration. With dementia cases rising globally and few disease-modifying treatments available, the idea that a routinely managed cardiovascular condition might hold a key to brain health is a prospect that will resonate far beyond the genetics community. The study, published as volume 27, article 126 of Biogerontology on 16 July 2026, received no dedicated funding and analyzed publicly available datasets from GTEx, FinnGen, the GWAS Catalog, the Million Veteran Program, and the UK Biobank-derived brain-age resource, ensuring that other researchers can readily build on its findings.