A phase 2a analysis moved six proteomic clocks in 42 IPF patients. The paper's own authors say the signal may reflect disease improvement, not slower aging. Here is how to read clock studies without getting fooled.
In plain terms A drug for lung disease made six age-prediction tests read younger in 42 patients over 12 weeks. The researchers say this could be because the drug helped the lungs, not because the patients got biologically younger. Diet and exercise studies show much smaller clock changes over years.
A secondary analysis of a 12-week phase 2a trial of rentosertib in 42 patients with idiopathic pulmonary fibrosis was published in Nature Biotechnology this month[1]. Rentosertib is an investigational drug for pulmonary fibrosis, not an anti-aging medicine.
The researchers applied six proteomic aging clocks to serum samples from trial participants: ProtAge, OrganAge, PAC, ipfP3GPT, PAOPAC, and a proteomic PhenoAge. All six indicated lower biological-age values in the treated arms than in placebo[1].
That is where the headlines usually stop. The paper does not stop there. The authors write that the clocks alone cannot separate an aging effect from an IPF-specific effect. That single sentence is the point of this article.
IPF is a progressive fibrotic lung disease that carries a poor prognosis. The disease drives inflammation and organ injury that show up in serum proteins. If a drug controls the disease, many of those proteins move. Aging clocks read the same proteins.
Swipe the chart sideways
| Signal source | Study type | Participants | Duration | What moved |
|---|---|---|---|---|
| Rentosertib in IPF | Phase 2a RCT, secondary analysis | 42 | 12 weeks | 6 proteomic clocks lower vs placebo; authors state IPF disease effect cannot be separated[1] |
| Caloric restriction | CALERIE, RCT, healthy adults | 220 | 2 years | DunedinPACE about 2-3% slower; Horvath and PhenoAge unchanged[4] |
| Clock-moving interventions | Systematic search of 41 human studies | ~4,000 | Weeks to years | Effects small and inconsistent across clocks and studies[8] |
| Cardiometabolic status | Cross-sectional, adults with obesity | n.r. | Snapshot | DunedinPACE tracked with cardiometabolic biomarkers; no intervention arm[10] |
| Complex disease burden | Review of epigenetic clocks and disease | Multiple cohorts | Cross-sectional | Horvath, PhenoAge, GrimAge accelerated in disease independent of chronological age[7] |
All references verified on PubMed on September 13, 2026.
Proteomic aging clocks are built on overlapping protein sets. Inflammation markers, growth factors, and injury-related proteins appear in more than one clock. If a treatment shifts these underlying proteins, all clocks trained on them will move in the same direction. That is not six independent confirmations. It is one signal read six ways.
A 2024 characterization of proteomic aging clocks noted that most were developed in midlife and late-life samples[5]. Extrapolating them to disease-modifying interventions was flagged as a limitation. The rentosertib analysis is exactly that kind of extrapolation.
A 2023 review of epigenetic clocks and complex disease showed that Horvath, PhenoAge, GrimAge, and other clocks all show accelerated values in complex disease independent of chronological age[7]. The same logic applies in reverse. Treat the disease, and the clock can move without touching biological aging.
The 2022 review titled Human age reversal: Fact or fiction? summarizes the field: aging clocks respond to disease, social variables, and mental health conditions[2]. A clock reading is not a direct measurement of biological age. It is a prediction from a training set. The prediction changes with anything in the training set that also changes.
CALERIE is the gold-standard human trial of caloric restriction. Healthy adults were randomized to 25% reduced calorie intake or ad libitum eating for two years. This is the largest and longest test of a lifestyle intervention with epigenetic clock endpoints.
DunedinPACE, a pace-of-aging clock, slowed by about 2 to 3 percent in the caloric restriction arm[4]. Horvath and PhenoAge did not change significantly. In a healthy population with a real intervention over two years, the effect on the clocks was small and inconsistent between clocks.
A systematic search published in 2026 identified 41 human studies of interventions that were reported to reverse epigenetic aging[8]. Effect sizes were small, and results were inconsistent across clocks and studies. The pattern is the same as CALERIE: modest movement, disagreement between clocks, no clear reversal.
Dietary geroscience methodology papers now recommend measuring multiple clocks and reporting agreement between them, along with sample size and timing considerations[9]. The methodological bar is rising because the field has learned that any single clock in isolation is easy to misread.
Swipe the chart sideways
A retrospective cohort study of DNA-methylation clocks and cardiovascular outcomes found that clocks correlate with incident cardiovascular events[6]. Their incremental value over standard cardiovascular risk factors was modest. That is useful research, but the direction runs from clock to outcome, not from intervention to clock to biology.
In adults with obesity, DunedinPACE tracked with cardiometabolic biomarkers[10]. The study was cross-sectional. It did not test whether treating cardiometabolic disease reversed the clock. It also did not test whether the clock added anything beyond the biomarkers a clinician would already measure.
The Cell 2023 framework on biomarkers of aging is explicit: any single biomarker in isolation is insufficient to declare an intervention geroprotective[3]. The framework calls for pairing molecular clocks with functional measures such as gait speed, VO2max, grip strength, and cognitive testing, alongside independent molecular readouts.
For an intervention trial to make a credible aging claim, movement on a clock is a starting point. What comes next is replication, functional outcomes, disease-specific controls, and a mechanistic story that ties the intervention to biology rather than to a prediction.
A drug moving six clocks in a phase 2a trial of 42 sick patients over 12 weeks is not a reason to buy anything. Rentosertib is investigational for IPF. It is not sold as a longevity drug. Any product that cites this paper as evidence for consumer anti-aging use is misusing it.
Home aging-clock tests will show a number. That number will move over time. Acute illness, stress, sleep, weight change, and normal lab variability can all shift the reading. Attributing the movement to a supplement or protocol requires the same rigor as any drug trial, and consumer tests do not run trials.
Interventions with functional endpoints beat interventions with only clock endpoints. Exercise, sleep quality, blood pressure control, and cardiometabolic management have decades of hard-outcome trials behind them, with effect sizes larger than anything currently measured on epigenetic clocks in lifestyle trials.
The right way to read this week's paper is the way the authors framed it. Aging biomarkers can be embedded in disease trials without waiting for mortality endpoints. That is a useful trial-design idea. It is not evidence that any human intervention has reversed aging.
Example: DunedinPACE in CALERIE
Interpret: hypothesis-generating
A single clock changing is a starting point, not a conclusion. Ask whether the change replicates in the other clocks measured in the same trial.
Signal onlyRead the CALERIE clock analysisLook for: walking speed, VO2max, cognition
Interpret: converging evidence
Aging clocks are more credible when paired with functional outcomes and independent molecular measures. A framework paper sets the bar for what counts as evidence.
MeaningfulRead the biomarker frameworkExample: rentosertib in IPF
Interpret: disease effect confounds
In a sick population, clock movement can reflect disease control. Even the study authors say the clocks alone cannot separate an aging effect from an IPF-specific effect.
Not proof of aging reversalRead the rentosertib paperRed flag: no functional endpoint
Red flag: no replication
Marketing that promises age reversal from one clock reading is not supported by the evidence. Human age reversal claims fail the tests laid out in the 2022 review.
Do not buyRead the review on age reversalA moving clock is a signal, not a verdict. Pair every clock with a functional outcome before spending money.
Six proteomic clocks all indicated lower biological-age values in the treated arm of a phase 2a trial in 42 patients with idiopathic pulmonary fibrosis over 12 weeks. The study authors explicitly state that the clocks alone cannot separate an aging effect from an IPF-specific effect. Rentosertib is investigational for IPF and is not an anti-aging treatment.
Proteomic clocks share overlapping proteins related to inflammation, fibrosis, and organ injury. A drug that reduces IPF disease activity can shift many of the same protein levels the clocks read. Correlated movement across clocks in a disease trial is not the same as evidence of aging reversal.
No. CALERIE tested 25% caloric restriction for two years in healthy adults. DunedinPACE slowed by about 2 to 3 percent. Horvath and PhenoAge did not change significantly. A comprehensive review found 41 human intervention studies with small and inconsistent effects across clocks.
Yes, in the right context. Clocks correlate with cardiovascular events, complex disease, and cardiometabolic status. Their incremental value over standard risk factors is modest. Consortium frameworks call for pairing clocks with functional outcomes and independent molecular measures before declaring an intervention geroprotective.
The result is a number that will move over time. It can be affected by acute illness, stress, sleep, weight change, and lab variability. No consumer-facing intervention has been shown to reverse aging on the basis of a home clock reading. Spend on interventions with functional endpoints first: exercise, blood pressure control, sleep, and diet.
Corrections and source queries: info@no1gevity.com.