New UT San Antonio study introduces analytical tool to better hone aging interventions
When scientists test potential life‑extending interventions, often the central question simply asks if the intervention works. New research from The University of Texas at San Antonio (UT San Antonio) suggests that the more meaningful questions may be when does it work and for whom?
In a study published in November’s issue of Nature Communications, scientists analyzed decades of data from the National Institute on Aging’s Intervention Testing Program. They found that many longevity‑promoting compounds do not have uniform effects across the lifespan. Instead, the benefits of these compounds, and in some cases their harms, can emerge at specific stages of life.
The work brings together expertise in aging biology and biostatistics and introduces a new analytical approach called the Temporal Efficacy Profiler. The tool — developed by a collaborative team led by James F. Nelson, PhD, professor in the Department of Cellular and Integrative Physiology at the Joe R. and Teresa Lozano Long School of Medicine and the Sam and Ann Barshop Institute for Longevity and Aging Studies, and Jonathan Gelfond, MD, PhD, vice chair and chief of the Division of Biostatistics in the Department of Population Health Sciences at the Long School of Medicine — allows researchers to pinpoint when an intervention is beneficial, potentially harmful or merely ineffective.
“This is really an example of how important good teams are to achieving scientific objectives,” said Nelson, a longtime investigator in the biology of aging and a key contributor to the Interventions Testing Program. “This paper would not have happened without that collaboration, especially the contributions of our students.”
A question of timing
For years, researchers studying aging interventions have relied on statistical methods that assume a treatment’s effect remains constant over time. The most common of these methods, known as the log‑rank test, is well-suited to answering binary questions like whether a treatment extends lifespan or not.
“But aging biology doesn’t work in binary terms,” Gelfond said. “An intervention might be helpful early in life, do nothing in midlife and actually become harmful later on. If you assume the effect is constant, you can miss all of that.”
The Temporal Efficacy Profiler was designed to address this limitation. Rather than evaluating survival outcomes as a single, averaged effect, the tool examines treatment effects day by day across the lifespan using a sliding analytical window. The method automatically adjusts the size of that window to balance precision and stability, allowing data, rather than researcher assumptions, to drive the results.
“What we wanted was a way to get precise estimates across time without imposing our own opinions,” Gelfond said. “The method essentially self‑tunes, which makes it more objective and scalable across many different interventions.”
The result is a visual and analytical profile that shows when a compound helps, when it has no effect and when it may actually pose risks.
Unexpected patterns and hidden harms
Using the profiler, the team analyzed lifespan data from 42 compounds tested in animal models through the Interventions Testing Program. Some of the findings confirmed previous work.
Well‑known compounds such as rapamycin and acarbose showed benefits across a broad range of ages, but others told a more complicated story.
Green tea extract, for example, appeared beneficial during certain periods of life but potentially harmful at others. This pattern was missed by conventional analytical tools because the opposing effects effectively canceled each other out.
“The standard log‑rank test is completely blind to that,” Gelfond said. “It just asks, does it help or hurt overall?”
The new analysis also revealed that several compounds had potential negative effects that had gone largely unrecognized.
“With this method, we identified roughly a dozen drugs that appear to have detrimental effects on survival,” Nelson said. “That simply hadn’t been visible before.”
These findings do not mean that commonly used supplements or interventions should be abandoned outright, the researchers emphasized. Instead, they highlight the importance of timing, dosage and careful follow‑up studies.
“This is not an end‑all, be‑all tool, it’s a way to point you in directions that need to be confirmed. Just like any result, you want replication in different populations and settings,” Nelson said.
Patterns of survival across the lifespan
The study builds on years of earlier work by the team and others examining how mortality risk changes with age. Previous analyses of large animal data sets revealed that risk is not constant over the lifespan. It tends to be higher in infancy, decreasing until puberty and then increasing exponentially thereafter. This pattern is largely mirrored in human populations.
“What stood out in those earlier studies was the consistency in the timing of these risk changes. Across species, you see similar age-related patterns in survival,” Nelson said.
When the researchers applied the profiler to intervention data, they found that the timing of when an intervention was given mattered as much as the intervention itself. Some treatments showed benefits primarily earlier in life, while others appeared to have stronger effects later in life. Why these shifts occur remains an open question, but the results suggest that age-specific responses are a critical factor in evaluating longevity interventions.
“If we can understand when an intervention is most likely to be beneficial, we can design strategies that are more precise and potentially more effective,” Nelson said.
From one‑size‑fits‑all to refined strategies
Beyond individual compounds, the researchers see broader implications for how longevity interventions might one day be deployed. Rather than administering a single drug continuously from early adulthood onward, future strategies could involve combining interventions targeted to specific life stages.
“Imagine using one compound early in life and a different one later, based on when each is actually effective,” Nelson said. “That could save time, reduce side effects and improve quality of life.”
Such ideas remain speculative, but the analytical framework now exists to explore them. The approach could also extend beyond aging research, Gelfond said, to areas such as cancer treatment, where therapies may offer strong short‑term benefits but diminishing or harmful effects over time.
“If patients and physicians had a clearer picture of when a treatment helps most, that could inform better decision‑making,” he said.
Informing public expectations
As interest in anti‑aging supplements and therapies continues to grow, the scientists urge caution when considering pharmaceutical longevity interventions.
“People often assume that if something is natural or popular, it must be safe and beneficial. But timing and dose matter, and there can be real downsides,” Nelson said.
Gelfond said the work highlights the importance of realistic expectations.
“If an intervention provides only a modest benefit, or only after years of use, people deserve to know that,” he said. “It helps them weigh cost, effort and potential side effects.”
A student‑driven collaboration
Both scientists emphasized that the study’s innovations were fueled by their students who ask questions that challenge conventional thinking.
“The students were really the key,” Gelfond said. “They asked big questions and didn’t accept the old all‑or‑nothing way of thinking.”
As the field of aging research evolves, the team hopes their work will encourage scientists to look beyond whether an intervention works and instead ask when, how and for whom it works best. In longevity science, they believe, timing may be everything.