370 Existing Drugs Could Target Aging, Network Biology Study Finds

Developing a drug from scratch for each of aging's root causes could take a decade or more per target. A team at Northeastern University, working with colleagues at Harvard, asked a different question: what if the drugs we already have are already sitting in the right molecular neighborhoods? Their answer, published in Nature Aging, is a computational map that flags 370 approved drugs as candidates to affect one or more of aging's eleven recognized hallmarks, though every prediction still requires experimental confirmation before anyone should update their medicine cabinet.

370 Approved Drugs Found Near Aging Hallmark Modules

The core finding is that aging-related genes are not scattered randomly across the human interactome; they cluster into specific, non-random modules. Albert-László Barabási's team identified 1,250 genes associated with 11 hallmarks of aging, ranging from epigenetic alterations to mitochondrial dysfunction, using the Open Genes database. They then mapped those genes onto the human interactome, a catalog of more than half a million protein-protein interactions. Because a drug does not need to hit an aging gene directly to influence its module, only to land close enough in the network to propagate an effect, the search space expanded dramatically. Running 6,442 drugs from the DrugBank database through that framework, the team flagged 370 with plausible proximity to at least one hallmark module. Of those, 14 carried a signal suggesting they could promote longevity; 14 others showed the opposite, an age-accelerating pattern worth monitoring. This kind of network medicine approach to identifying aging drugs has been proposed before in narrower contexts, but this study applies it at a scale that covers most of the approved pharmacopeia.

Aspirin and a Nasal Spray Decongestant Among the Candidates

Two drugs illustrate how counterintuitive the candidates can be. Aspirin, already prescribed in low doses for cardiovascular protection and currently being tested in human longevity trials, showed up as positively affecting two hallmarks: cell-cell communication and deregulated nutrient sensing. That alignment with existing human-trial interest helps validate the method. More surprising is oxymetazoline, the active ingredient in over-the-counter nasal sprays like Afrin and Sinex, as well as some eye drops and topical creams. The network analysis flagged it as a candidate to improve cell-cell communication, a hallmark whose deterioration is linked to tissue aging. Harvard collaborators are now running cell-line experiments to test that prediction directly. As Barabási put it, "The challenge is figuring out which drugs are worth testing." The network approach is meant to answer exactly that, by narrowing a field of thousands down to a tractable list.

How Mapping Protein Interactions Made the Shortlist Credible

The credibility of the shortlist rests on the team's validation step. Before projecting the 6,442-drug library, the researchers confirmed that their framework could recover 17 drugs already enrolled in human longevity trials and 11 compounds known to extend lifespan in mice. Hitting those benchmarks suggests the metric is capturing a real signal rather than noise. Bnaya Gross described the organizing logic: "Networks allow us to organize them, saying, 'OK, it's not just a pile of genes. They are connected to each other. They form some sort of organization. It's not a random process.'" That organizational insight is what separates this approach from earlier lists of candidate longevity drugs, which were often assembled from disparate experimental results without a unifying framework. The advantages of repurposing FDA-approved drugs are well established: known safety profiles and existing manufacturing infrastructure mean that a positive experimental signal can move toward clinical testing faster than a novel compound.

The Gap Between Computational Prediction and Human Evidence

Barabási was direct about what the study does and does not show: "This study does not provide a cure for aging, nor does it prove that any specific drug will extend human life. It offers a roadmap toward actionable interventions that can be tested in cells, animals and eventually humans." That roadmap framing is the right one. The 370 flagged drugs are hypotheses, not treatments. Each prediction needs to survive cell-line testing, then animal models, then human trials before it tells us anything clinically actionable. Translating anti-aging drug findings from bench to bedside has historically been slow; many compounds that extend lifespan in model organisms have produced inconclusive results when tested in people. The complexity compounds because aging is not a single process: the science involves genetics, environment, and dozens of interacting molecular pathways. The researchers acknowledge that addressing multiple hallmarks would likely require several different interventions working in combination, not a single repurposed pill.

Precise Geroscience as the Longer-Term Goal

The team's ambition extends beyond generating a shortlist. They envision the methodology eventually enabling what they call "precise geroscience," tailoring anti-aging interventions for specific individuals based on which hallmarks are most active in their biology at a given time. That would require integrating patient-level molecular data with the network framework, a step that remains future work. In the nearer term, the oxymetazoline cell-line experiments represent the first experimental test of a prediction the model generated independently rather than retrospectively. If those results hold, they would provide direct evidence that the network proximity metric is identifying genuine biological relationships and not just statistical coincidence. The broader implication is that a large fraction of aging biology may already be druggable with compounds sitting on pharmacy shelves, waiting to be matched to the right molecular context.

Evidence rating

INVESTIGATIONAL — 61/100

Caveats:

• This is a computational, network-based study — no human or animal lifespan data were generated.
• Predictions have not yet been validated in clinical trials; cell-line experiments are still ongoing.
• Funding source was not disclosed by the authors.

Computational, network biology study mapping 1,250 aging-related genes onto the human protein interactome to predict drug repurposing candidates. No human or animal lifespan data were generated; findings are model-derived predictions requiring cell, animal, and clinical validation. Not yet replicated independently.

References

1. Bnaya Gross, Joseph Ehlert, Vadim N. Gladyshev, Joseph Loscalzo, Albert-László Barabási (2026). Network-driven discovery of repurposable drugs targeting hallmarks of aging. Nature Aging. https://doi.org/10.1038/s43587-026-01161-8
2. Terry W. Snell, Rachel K. Johnston, Bharath Srinivasan, Hongyi Zhou, Mu Gao, Jeffrey Skolnick (2016). Repurposing FDA-approved drugs for anti-aging therapies. Biogerontology. https://doi.org/10.1007/s10522-016-9660-x
3. Alexander Vaiserman, Alexander Koliada, Oleh Lushchak, Manuel J. Castillo (2020). Repurposing drugs to fight aging: The difficult path from bench to bedside. Medicinal Research Reviews. https://doi.org/10.1002/med.21773