Can the Oldest Galactic Stars Reveal the True Age of the Universe?

Key Takeaways

• A cleaned sample of 155,600 stars places the oldest stellar age near 13.7 billion years.
• The result agrees with CMB-based ΛCDM and challenges cosmologies that shorten cosmic history.
• Shared uncertainties in stellar models prevent the study from delivering a final cosmological verdict.

A Stellar Census Points to a 13.7-Billion-Year Clock

On July 1, 2026, Indranil Banik, Thenujaya Kudakolawa Kaluarachchige, Stephen Cookson, and Harry Desmond posted a new stellar-age study based on a large population of ancient Milky Way stars. Their central result places the latent age of the oldest star in the sample at 13.73 billion years, with a statistical interval of plus 0.18 billion years and minus 0.15 billion years.

Adding an assumed 0.2 billion years between the Big Bang and the formation of long-lived stars produces an estimated cosmic age near 13.93 billion years. That estimate is compatible with the roughly 13.8-billion-year age obtained from cosmic microwave background measurements under the standard Lambda cold dark matter model, abbreviated ΛCDM.

The analysis uses 155,600 selected stars within 5 kiloparsecs, or about 16,300 light-years, of the Sun. The starting sample contained 247,103 stars after the authors removed a duplicate entry from the original catalogue. Its measurements combine spectroscopy from the LAMOST DR7 survey with astrometry from the European Space Agency’s Gaia mission.

Gaia collected more than 3 trillion observations of about 2 billion stars and other objects between July 27, 2014, and January 15, 2025. Its measurements of stellar positions, motions, luminosities, temperatures, and compositions have changed how astronomers reconstruct the formation of the Milky Way. The 2026 analysis uses Gaia early Data Release 3 parallaxes, supplemented by stellar information from Gaia Data Release 3.

The stars in the sample are mainly subgiants, objects that have begun leaving the stable hydrogen-burning stage of their lives. Their temperatures and luminosities change relatively quickly during this phase. Stellar models can often estimate a subgiant’s age more precisely than the age of a comparable main-sequence star.

The result has attracted attention because it turns nearby stars into an independent test of cosmology. The work remained an arXiv preprint as of July 18, 2026. Its manuscript uses the Monthly Notices of the Royal Astronomical Society format, but the acceptance and receipt fields still contain placeholders. The methods and conclusions should be treated as provisional until journal review, replication, and broader testing are complete.

The study belongs within the longer effort to determine how old the Universe is through independent clocks. Cosmic microwave background radiation, expansion measurements, radioactive dating, globular clusters, stellar populations, and early galaxies probe different parts of cosmic history. Agreement among these methods strengthens confidence in the chronology. Disagreement can expose measurement errors, incomplete astrophysics, or weaknesses in the cosmological model.

How Ancient Stars Place a Lower Bound on Cosmic Time

No star can be older than the Universe that contains it. This ordering turns stellar ages into a lower bound on cosmic age, but the calculation is far from simple. Astronomers must estimate a star’s mass, chemical composition, distance, luminosity, temperature, and evolutionary stage, then compare those observations with theoretical tracks describing how stars change over time.

The Banik team used age likelihoods derived from the Xiang and Rix catalogue. That earlier study combined Gaia astrometry, LAMOST spectroscopy, Gaia photometry, and Two Micron All Sky Survey measurements. It reported a median relative age precision of about 7.5% for its subgiant sample and used Bayesian matching to Yonsei-Yale stellar isochrones.

An isochrone is a theoretical curve representing stars of the same age and chemical composition but different masses. A star’s observed temperature, luminosity, and composition are compared with these curves to obtain a probability distribution for its age. This approach produces a range of plausible ages rather than a single perfectly known value.

The 2026 paper preserves the complete age likelihood for each star on a grid extending to 20 billion years. That unusually high ceiling reduces the chance of building the accepted cosmological age directly into the stellar analysis. A ceiling fixed near 13.8 billion years would prevent the data from producing older apparent ages, even when measurement errors or model weaknesses warranted them.

Apparent ages above the true age of the Universe are expected in a sufficiently large catalogue. Measurement uncertainty can push a star’s estimated age above its actual age. The scientific problem is not whether any individual star receives an age above 13.8 billion years. The problem is determining where the underlying population ends after accounting for those errors.

The stellar clock also requires a formation interval. Long-lived stars did not appear at the instant of the Big Bang, so the age of the oldest surviving star must be lower than the age of the Universe. The authors adopt a delay of 0.2 billion years, consistent with models and observations indicating that star formation began within the opening few hundred million years of cosmic history.

Observations by the James Webb Space Telescope have revealed galaxies at very high redshifts, confirming that organized star formation began early. These detections do not prove that a surviving Milky Way star formed precisely 0.2 billion years after the Big Bang. They support a relatively short formation delay rather than a gap lasting 1 billion years or more.

The distinction means that the study does not measure cosmic age in the same manner as the cosmic microwave background. Its most direct result concerns the oldest-star cutoff, represented as A-star in the paper. Converting that cutoff into the age of the Universe requires an astrophysical estimate for the formation delay.

A longer delay raises the inferred cosmic age and increases the disagreement with younger cosmological models. A shorter delay reduces the total, though the interval cannot become negative. The formation delay becomes more significant once the statistical uncertainty in the stellar cutoff falls below a few hundred million years.

Nearby stellar fossils provide a complementary method to deep-field cosmology. Telescopes observing distant galaxies look backward through light-travel time. Galactic archaeology reconstructs history from stars that survived until the present. New Space Economy’s introduction to modern cosmology explains why independent methods matter: each relies on different observations, calculations, and assumptions.

Building a Trustworthy Sample from Gaia and LAMOST

A large catalogue does not automatically produce an accurate age boundary. A small fraction of badly measured stars can distort an attempt to locate the oldest member of a population. Some objects in the starting catalogue appeared older than 16 billion years and had unexpectedly small age uncertainties.

That pattern was physically suspicious. Very old stars are faint and difficult to characterize, and objects extending far beyond the real age boundary should generally owe their extreme values to substantial errors. The paper estimates that about 0.1% of the initial sample had problematic age determinations.

The team imposed a parallax requirement that restricted the sample to stars within 5 kiloparsecs. Each accepted star also needed a parallax uncertainty below 10%. Reliable parallax measurements help astronomers convert apparent brightness into absolute luminosity, an important input for age estimation.

Possible unresolved multiple-star systems were removed. Light from a companion can make an object appear brighter than a single-star model expects, leading to an incorrect estimate of luminosity, mass, or age. The authors used a Gaia image-processing parameter that records how often a source appeared as more than one peak during focal-plane transits.

The more distinctive selection rules involve stellar chemistry. Ancient stars formed before many generations of stellar explosions enriched the Galaxy with heavy elements. They should usually have low iron abundance, expressed through the quantity [Fe/H], and elevated alpha-element abundance relative to iron, expressed as [alpha/Fe].

Core-collapse supernovae produced alpha elements early in Galactic history. Type Ia supernovae supplied a larger share of iron after a delay. An apparently 17-billion-year-old star with near-solar metallicity and little alpha enhancement is more likely to have an incorrect age estimate than to overturn the standard cosmic chronology.

The authors constructed age-dependent chemical boundaries and removed stars occupying sparsely populated, physically doubtful regions of the age-metallicity and age-alpha diagrams. The chart on page 4 of the paper shows a downward-sloping metallicity boundary. Older accepted stars must become progressively more metal-poor. The adjacent alpha-enhancement chart on page 5 requires increasingly old stars to have chemistry consistent with early formation.

A further cross-check used ages from Gaia’s Final Luminosity Age Mass Estimator, known as FLAME. The FLAME processing module derives stellar masses and evolutionary properties using Gaia data and stellar models.

Published FLAME ages had an artificial ceiling of 13.5 billion years, with older values omitted. Excluding every star without a published FLAME age would have biased the sample against the oldest candidates. The team instead used the tight relationship between subgiant age and mass to estimate FLAME ages for 3,302 stars that retained mass estimates but lacked published ages.

Only 455 stars with imputed FLAME ages remained in the nominal sample. The authors then compared FLAME ages with the more information-rich Yonsei-Yale ages based on LAMOST and Gaia. Stars lying far from the population trend were removed.

After all cuts, 155,600 stars remained. Their age uncertainty rose continuously toward older ages rather than falling at the extreme end. The authors treat that pattern as evidence that the suspicious high-age tail had been cleaned.

The sensitivity tests show how sample selection affects the answer. Raising the age-dependent metallicity ceiling increased the inferred oldest-star age to 14.02 billion years, but it began to reintroduce questionable uncertainty behavior. Lowering the ceiling by 0.2 dex reduced the estimate to 13.31 billion years, yet that restrictive cut passed through a densely populated stellar sequence and removed 13,371 stars.

The nominal result of 13.73 billion years sits between these extremes. The sample-selection method is one of the paper’s strongest features, but it also involves scientific judgment. Chemical expectations have a firm astrophysical basis, though the exact cutoff lines were selected manually.

The authors state that the quality cuts were defined before important stages of the oldest-age inference were developed. That sequencing reduces the chance that the boundaries were adjusted to recover a preferred cosmological result. Independent replication with other surveys, stellar models, and selection algorithms would provide a stronger test.

Reconstructing the Hidden Age Distribution

Observed ages are blurred versions of true ages. In a catalogue containing more than 150,000 stars, some measurements will sit several standard deviations above their real values even when the error model works properly. Selecting the largest reported age would almost guarantee an overestimate.

The paper addresses this extreme-value problem through several statistical methods. It begins with intuitive checks involving the oldest individual likelihoods and ends with a Markov Chain Monte Carlo reconstruction of the full population.

The simpler methods test how many stars have age likelihoods extending beyond a proposed cutoff. If the true oldest age were only 12.5 or 13 billion years, many stars in the cleaned sample would appear improbably older than the boundary. Cutoffs near 13.6 to 13.7 billion years produce behavior closer to the number of expected statistical outliers.

The authors also constructed simulated catalogues with assumed stellar-age distributions. They added Gaussian measurement errors of 7.5% or 10% and repeated the procedure 20,000 times. These tests placed the expected oldest-star cutoff in approximately the same range as the later reconstruction.

The main calculation estimates the latent age distribution, meaning the distribution of true ages before observational errors distort it. The authors divide the age interval into 100 bins and use a Markov Chain Monte Carlo method to search for population distributions that overlap with every star’s individual likelihood.

A smoothing penalty discourages unsupported fluctuations between adjacent bins. The paper also tests a scale-invariant prior and a version with no smoothing penalty. Removing the penalty makes the reconstructed distribution noisier but changes the inferred cutoff by only about 0.14 billion years.

The reconstruction cannot return negative probabilities. It consequently leaves a small positive numerical floor beyond the real stellar boundary. The authors divide the reconstructed probability in each age bin by its uncertainty to distinguish a supported stellar signal from this numerical floor.

Between roughly 12 and 14 billion years, the probability-to-uncertainty ratio falls almost linearly. At higher ages it becomes approximately flat. A line-plus-flat model identifies the transition between these regimes as the oldest-star boundary.

That procedure yields 13.73 billion years, with a statistical interval of plus 0.18 billion years and minus 0.15 billion years. Removing the smoothing penalty gives a result near 13.59 billion years. Changing the prior produces a value near 13.76 billion years.

The spread among these analysis choices is similar in size to the formal statistical uncertainty. It suggests that the reconstruction method contributes an additional uncertainty of roughly one or two tenths of a billion years, even before shared stellar-model errors are considered.

A smaller chemical sample provides another internal comparison. The authors selected stars associated with Gaia-Sausage-Enceladus, the remains of a galaxy that merged with the Milky Way early in its history. Their chemistry and origin imply a narrower formation-age distribution.

A simpler parametric model applied to this much smaller population produced an oldest age close to the main result. The formal error was narrower, but the paper cautions that the assumed distribution may not describe the true population perfectly. Agreement in the central age is more informative than the small fitted error.

The method is a thoughtful response to the statistical problems created by a large catalogue. It also introduces an estimator that requires outside testing. Simulated catalogues can show whether the method recovers a known cutoff under selected assumptions, but real stellar populations may contain correlated errors, unexpected chemical groups, or selection effects that the simulations do not capture.

Independent teams can test whether the probability-to-uncertainty crossover remains stable under different priors, error models, age grids, and stellar-evolution calculations. The public arXiv manuscript provides a detailed mathematical description of the reconstruction.

Agreement With the Cosmic Microwave Background Age

The cosmic microwave background is relic radiation released when the Universe became transparent about 380,000 years after the Big Bang. Its temperature and polarization patterns encode the densities of matter, radiation, and other cosmological components.

Under flat ΛCDM, these measurements determine an expansion history and produce a cosmic age near 13.8 billion years. The SPT-3G D1 analysis found cosmological parameters consistent with results from Planck and the Atacama Cosmology Telescope.

SPT-3G reported a Hubble constant of 66.66 ± 0.60 kilometers per second per megaparsec from its own data. Combining SPT-3G, Planck, and Atacama Cosmology Telescope measurements produced 67.24 ± 0.35 kilometers per second per megaparsec.

The stellar result fits this chronology closely. Cosmic microwave background-calibrated ΛCDM predicts a Universe about 13.8 billion years old. Subtracting a 0.2-billion-year formation delay gives an expected maximum stellar age near 13.6 billion years.

The nominal stellar estimate of 13.73 billion years is slightly higher but statistically compatible. Analysis choices and shared stellar-model uncertainties broaden the comparison further. A formation delay somewhat shorter than 0.2 billion years would also bring the values closer.

Agreement does not prove ΛCDM. The cosmic microwave background age depends on the cosmological model, and the stellar age depends on models of stellar evolution. The result shows that two methods using different observations and physical calculations occupy the same narrow time interval.

The microwave background calculation relies on early-universe plasma physics, general relativity, and a parameterized cosmological model. Stellar dating relies on nuclear burning, opacity, convection, chemical composition, distance, and photometry. Matching answers reduce the likelihood that either method contains an unidentified error large enough to shift cosmic age by approximately 1 billion years.

The convergence connects with New Space Economy’s account of why the accepted age changed. Stellar ages once created an age crisis because some stars appeared older than cosmological estimates. Improved distance measurements, stellar physics, and evidence for accelerated expansion removed that contradiction.

The new paper reverses the older logic. Ancient stars now test proposed cosmologies that would make the Universe substantially younger. A cosmological model that produces an age near 12.9 billion years must explain how long-lived stars could already be about 13.7 billion years old.

Early-galaxy observations provide context without directly fixing the total age. Webb has found organized galaxies only a few hundred million years after the Big Bang. New Space Economy’s examination of rapid early galaxy formation explains that these objects do not automatically require an older Universe.

They may instead require faster star formation, revised stellar-mass estimates, different dust assumptions, or improved galaxy-formation models. Such observations support the plausibility of a short stellar-formation delay, but they do not identify the birth date of the oldest surviving Milky Way subgiant.

Pressure on Purely Early-Time Hubble-Tension Solutions

The Hubble tension is the disagreement between the expansion rate inferred from early-universe observations and the rate measured through nearby distance indicators. Cosmic microwave background analyses under ΛCDM cluster near 67 kilometers per second per megaparsec.

The Local Distance Network combined Cepheids, red giant stars, masers, supernovae, surface-brightness fluctuations, and other distance methods. It reported a community consensus value of 73.50 ± 0.81 kilometers per second per megaparsec. Compared with the combined cosmic microwave background result, the difference is about 7.1 standard deviations.

Many proposed solutions alter physics before recombination so that the sound horizon becomes smaller. A smaller sound horizon can make cosmic microwave background observations compatible with a higher present expansion rate.

Proposals include early dark energy, modified recombination, changing particle properties, additional relativistic particles, and primordial magnetic fields. New Space Economy has examined primordial magnetic fields and the broader question of why expansion measurements disagree.

A cosmology with a higher expansion rate over much of cosmic history generally leaves less elapsed time between the Big Bang and the present. The Banik paper combines a local Hubble rate with a low-redshift matter-density estimate and obtains a representative cosmic age of 12.91 ± 0.18 billion years for models in which early-time physics resolves the tension and later evolution resembles ΛCDM.

Subtracting 0.2 billion years for stellar formation predicts a maximum stellar age of 12.71 ± 0.18 billion years. The nominal reconstructed stellar cutoff exceeds that prediction by about 1 billion years.

This comparison is the paper’s strongest cosmological claim. It does not disprove every model containing early-universe modifications. A model could alter the shape of the expansion history, change the relationship between low-redshift measurements, or modify assumptions used in stellar physics.

The result instead creates a demanding age test for proposals that reduce total cosmic history to roughly 12.9 billion years. Such models must explain how stars appear to have survived for about 13.7 billion years before any allowance is added for their formation.

Late-time or local explanations face less pressure from the stellar clock because they change expansion during a smaller part of cosmic history. A transition in dark energy at low redshift, an unidentified distance-calibration effect, or local density structure could alter nearby expansion measurements without removing 1 billion years from the earlier chronology.

This does not establish that any late-time explanation is correct. It shows that stellar age measurements can separate classes of Hubble-tension proposals according to when their new physics operates.

The result functions as a model-selection test rather than a complete solution. The stellar clock favors cosmologies that retain an old Universe, including CMB-calibrated ΛCDM and some low-redshift modifications. It weighs against early-only remedies that require substantially less cosmic time.

Other observations still test whether any alternative model fits the full cosmic microwave background spectrum, baryon acoustic oscillations, primordial element abundances, supernova distances, gravitational lensing, and structure growth.

Stellar Physics Sets the Main Limit on Certainty

The formal statistical error of roughly 0.17 billion years is smaller than the complete uncertainty. Stars are modeled through assumptions about convection, helium abundance, opacity, diffusion, nuclear reaction rates, element mixtures, mass loss, and atmospheric properties.

An error shared by the stellar models could shift every age in the same direction. Increasing the catalogue size would not remove that common offset.

The Yonsei-Yale tracks used in the source catalogue are older than some alternative isochrone systems. Xiang and Rix compared them with MESA Isochrones and Stellar Tracks and found that the alternative models produced ages about 0.5 billion years higher for some of the oldest stars when solar alpha-element abundance was assumed.

A modern comparison incorporating the correct alpha enhancement for each star could provide a stronger estimate of model dependence. The 2026 paper notes that newer MIST II models may offer such an opportunity once suitable calculations become available.

The authors devote attention to the mixing-length parameter, a simplified description of convection in one-dimensional stellar models. The Yonsei-Yale tracks use a value near the solar calibration. Metal-poor subgiants might behave differently from the Sun, but recent asteroseismic modeling of HD 140283 obtained a value relatively close to the solar result.

Initial helium abundance is another source of uncertainty. Higher helium content accelerates stellar evolution and reduces the inferred age for a star observed at a given evolutionary stage. The effect is connected to metallicity because later generations of stars contain more helium and heavy elements produced by earlier stars.

The oldest candidates in the sample are highly metal-poor. Their original helium abundance should consequently remain fairly close to the primordial value. The paper estimates that uncertainty in helium enrichment could shift the oldest-age result by roughly 0.15 billion years, though this estimate depends on stellar calculations and chemical-enrichment assumptions.

Independent work illustrates why caution remains necessary. A 2024 study using tailored abundances modeled HD 140283 at 12.3 billion years. Applying a solar-scaled chemical mixture to the same star produced an age near 14 billion years.

A subsequent asteroseismic analysis used oscillations measured by the Transiting Exoplanet Survey Satellite and obtained an age of 14.2 ± 0.4 billion years. The team combined oscillation frequencies with mass, radius, chemistry, and other constraints.

The difference between these estimates demonstrates how chemical mixtures, convection, opacities, and observational constraints can change a stellar age. Asteroseismology supplies direct information about a star’s interior, but it still requires stellar models.

The source catalogue’s individual errors also deserve attention. A median precision of 7.5% corresponds to about 1 billion years for a 13.7-billion-year star. The new analysis obtains a much tighter population cutoff by combining the complete age likelihoods of many stars.

This statistical gain is valid when the errors are correctly described and sufficiently independent. Correlated temperature scales, parallax offsets, abundance calibrations, isochrone assumptions, or selection effects can leave a narrow-looking population result that remains displaced as a group.

Preprint status matters for this reason. Journal review cannot guarantee correctness, but it can identify coding errors, underestimated systematics, unclear priors, or weaknesses in a new estimator. Public release of the selected catalogue, likelihood files, analysis code, and simulation tests would make replication easier.

The paper states that the original Xiang and Rix catalogue is public. Its selected-star list and likelihood products are available from the authors on reasonable request. A permanent public archive would make detailed independent testing more accessible.

Independent Stellar Clocks Support an Old Universe

The new estimate does not stand alone. HD 140283 has served as a benchmark because it is nearby, bright, metal-poor, and old. Its evolutionary stage makes its luminosity sensitive to age.

The 2025 asteroseismic estimate of 14.2 ± 0.4 billion years is statistically consistent with a 13.8-billion-year Universe. A single star cannot carry the population-level weight of 155,600 selected objects, but its oscillation data probe stellar interiors in a manner that complements luminosity-and-temperature fitting.

A separate Gaia stellar-age study selected 160 well-measured ancient stars from a larger catalogue. It found an age distribution centered near 13.6 billion years, with statistical and systematic uncertainties considerably larger than those claimed by the Banik paper.

Assuming a 0.2-billion-year formation delay, that study obtained a lower bound near 13.8 billion years for the age of the Universe. Its agreement with the larger LAMOST and Gaia analysis supports an old cosmic chronology, though the two studies use different samples and statistical approaches.

Globular clusters provide another clock. Their stars share a common distance, similar formation history, and related chemical composition. These common properties can simplify relative dating.

Globular clusters also contain multiple stellar populations and require careful treatment of distance, reddening, helium, stellar diffusion, and chemical enrichment. Modern cluster estimates generally remain compatible with a Universe close to 13.8 billion years old.

The persuasive feature is the clustering of independent estimates above 13 billion years. Field subgiants, globular clusters, asteroseismology, and cosmic microwave background cosmology use different observations and face different uncertainties. Their overlap makes a 12.5-to-12.9-billion-year Universe difficult to accommodate without coordinated shifts across more than one method.

Future data can sharpen the test. Gaia Data Release 4 was expected in December 2026 as of July 18, 2026, subject to processing and validation. It is planned to include longer observational baselines, additional source classifications, and expanded stellar products.

The Vera C. Rubin Observatory began its 10-year Legacy Survey of Space and Time on June 30, 2026. Its repeated imaging will identify ancient halo populations, variable stars, binary systems, stellar streams, and merger debris for follow-up.

Rubin does not replace detailed spectroscopy for precise chemical dating. Its survey can improve target selection, variability screening, binary identification, and the mapping of ancient Galactic structures. Spectroscopic observatories can then measure the elemental abundances needed for age estimates.

Asteroseismology may provide the strongest improvements for selected ancient stars. Oscillation frequencies reveal mass, density, internal structure, and convection. Combining those data with star-specific abundances and Gaia distances can test whether population isochrone ages share a common offset.

Wide binaries offer another calibration route. Stars in a binary generally formed together and should return compatible ages despite having different masses. They cannot reveal every shared model error, but they can test internal consistency across different evolutionary states.

Better estimates of the formation delay are also needed. Webb observations can push the record of early galaxies deeper into cosmic dawn. Connecting those galaxies to surviving low-mass stars requires models of metal enrichment, gas cooling, fragmentation, star formation, and Milky Way assembly.

A delay uncertainty of 0.1 or 0.2 billion years once appeared small beside billion-year stellar-age errors. It now matters because the reconstructed population cutoff has reached similar statistical precision.

What the Study Changes in the Cosmology Debate

The paper adds an independent age dimension to a dispute often presented as a contest between two Hubble-constant values. Expansion rate, matter density, dark-energy behavior, and cosmic age are connected through the history of expansion.

A model cannot raise the present Hubble constant freely without consequences for elapsed time, distances, structure formation, and early-universe observables. The oldest Galactic stars require proposed solutions to leave enough time between the Big Bang and the present.

This framing helps distinguish measurement tension from model tension. Local distance measurements could be accurate, cosmic microwave background measurements could be accurate, and ΛCDM could still fail to connect them.

A replacement model must fit more than those two datasets. It must preserve enough cosmic time for ancient stars, match primordial element abundances, reproduce baryon acoustic oscillations, and remain compatible with galaxy formation and structure growth.

New Space Economy’s survey of cosmological paradoxes places the Hubble tension among several unresolved tests of modern cosmology. These problems need not share a single cause.

The preprint strengthens the case against solutions that maintain a faster expansion rate across most of cosmic history. It provides less direct guidance about low-redshift dark energy, local structure, distance-calibration errors, or models with more complicated expansion histories.

The authors favor a low-redshift explanation if their stellar result survives. A neutral assessment permits several outcomes. The stellar age scale could move after model revisions, the local expansion value could shift after calibration work, or cosmology could require a model that raises the locally inferred Hubble constant without making the Universe too young.

One numerical result should not be treated as a verdict. The 13.73-billion-year cutoff depends on manually selected chemical boundaries, a new statistical crossover method, Yonsei-Yale isochrones, assumed likelihoods, and the treatment of the high-age numerical floor.

The sensitivity tests mostly retain an old Universe. Even the restrictive metallicity selection produces an oldest-star estimate of 13.31 billion years before a formation interval is added. That result remains difficult to reconcile with a cosmic age of 12.91 billion years.

Independent replication would make the age test more influential. Other teams could begin with Gaia and LAMOST, substitute different isochrone families, revise abundance scales, model correlated errors, and use hierarchical statistics tied to a physical star-formation history.

Agreement near 13.7 billion years would turn ancient stars into a powerful filter for cosmological proposals. A substantial downward shift would identify where stellar modeling, sample selection, or the cutoff estimator had overstated the age.

Nearby stars can be measured through detailed spectra and precise astrometry, providing direct access to chemistry and evolutionary state. Distant-universe observations probe expansion and primordial conditions. When both methods point toward the same cosmic age, proposed solutions to the Hubble tension must account for that agreement.

Summary

The July 2026 preprint derives an oldest-star cutoff of 13.73 billion years from 155,600 selected Milky Way subgiants. Its sample combines LAMOST spectroscopy with Gaia astrometry, then uses chemical expectations, binary screening, parallax quality, and cross-catalogue age agreement to remove suspicious cases.

A Markov Chain Monte Carlo reconstruction estimates the hidden population-age distribution. The method identifies the transition between a declining stellar signal and the numerical probability floor as the oldest-star boundary.

Adding a 0.2-billion-year formation delay gives a cosmic age near 13.9 billion years, compatible with the standard 13.8-billion-year result from cosmic microwave background cosmology. The result places pressure on early-time solutions to the Hubble tension that imply a Universe near 12.9 billion years old.

It does not eliminate every early-universe modification or identify the source of the Hubble tension. The largest unresolved issue is shared stellar-model uncertainty. Helium abundance, convection, chemical mixtures, temperature scales, parallax calibration, and isochrone physics can move many stellar ages together.

The paper’s preprint status and new statistical estimator call for independent replication. Even with those cautions, the overlap among field stars, globular clusters, asteroseismology, and cosmic microwave background measurements supports a Universe close to 13.8 billion years old.

Appendix: Useful Books Available on Amazon

• A Brief History of Time
• The Fabric of the Cosmos
• The Little Book of Cosmology
• The Cosmic Cocktail
• The End of Everything

Appendix: Top Questions Answered in This Article

What Age Did the Study Estimate for the Oldest Star?

The nominal estimate is 13.73 billion years, with a statistical interval of plus 0.18 billion years and minus 0.15 billion years. That value describes the inferred cutoff of the underlying stellar-age distribution, rather than the largest reported age assigned to one observed star.

How Does the Stellar Result Translate Into the Age of the Universe?

The authors add an estimated 0.2 billion years for the interval between the Big Bang and the formation of long-lived stars. This produces a cosmic age near 13.93 billion years. The formation interval remains an astrophysical estimate rather than a directly measured constant.

Why Did the Researchers Use Subgiant Stars?

Subgiants change temperature and luminosity relatively quickly after leaving stable hydrogen burning. Their positions on a temperature-luminosity diagram are more sensitive to age than those of many main-sequence stars. Precise distance and chemical measurements consequently make subgiants useful stellar clocks.

Why Were So Many Stars Removed From the Starting Catalogue?

The team removed stars with weak parallaxes, possible unresolved companions, chemically implausible old ages, or substantial disagreement between two age-estimation systems. These cuts reduced the sample from 247,103 stars to 155,600. The purpose was to prevent a small number of unreliable ages from controlling the population boundary.

Does the Result Prove That the Universe Is 13.8 Billion Years Old?

No. It provides an independent stellar estimate compatible with that age under standard stellar physics and a short formation delay. Shared model errors could shift the stellar scale, and the cosmic microwave background age also depends on the selected cosmological model.

What Does the Study Imply for the Hubble Tension?

It places pressure on solutions that increase the expansion rate through much of cosmic history and reduce the cosmic age to about 12.9 billion years. Such models leave too little time for stars inferred to be about 13.7 billion years old. Low-redshift or local explanations are less affected by this age test.

Can Measurement Errors Make Stars Appear Older Than the Universe?

Yes. Large catalogues naturally contain statistical outliers, and stellar-model errors can produce ages above the real cosmic boundary. The study avoids selecting the largest observed age and instead reconstructs the underlying population from each star’s complete age likelihood.

What Is the Largest Uncertainty in the Result?

Shared stellar-model systematics are more limiting than the formal sample statistics. Helium abundance, convection, chemical mixtures, temperature calibration, opacity, and isochrone assumptions can shift many ages together. Increasing the number of stars does not automatically remove a common modeling offset.

Why Is HD 140283 Relevant?

HD 140283 is a nearby, metal-poor subgiant used to test stellar-evolution models. A 2025 asteroseismic study estimated its age at 14.2 ± 0.4 billion years. Other model assumptions have produced lower estimates, demonstrating how strongly stellar chemistry and interior physics can affect age calculations.

What Observations Could Improve the Stellar Age Test?

Gaia Data Release 4, Rubin Observatory measurements, new spectroscopic surveys, and more asteroseismic observations can improve distances, chemistry, binary identification, and stellar masses. Public release of the selected sample and analysis products would also allow independent researchers to test the statistical method and model assumptions.

Appendix: Glossary of Key Terms

Age Likelihood

A probability distribution describing which ages are compatible with a star’s measured properties and an adopted stellar model. It preserves uncertainty and asymmetry that would be lost if the result were reduced to one age and one error bar.

Alpha Elements

Elements such as oxygen, magnesium, silicon, sulfur, calcium, and titanium that are produced heavily by core-collapse supernovae. Their abundance relative to iron helps astronomers identify stars formed early in the Milky Way’s chemical history.

Asteroseismology

The study of oscillations inside stars. Measured vibration frequencies constrain stellar density, mass, radius, internal structure, and convection, allowing age estimates that complement methods based mainly on luminosity, temperature, and chemical composition.

Cosmic Microwave Background

Relic radiation released when the early Universe cooled enough for atoms to form and light to travel freely. Its temperature and polarization patterns constrain cosmological parameters, including the expansion rate, matter density, and model-dependent cosmic age.

FLAME

The Final Luminosity Age Mass Estimator used in Gaia data processing. It derives stellar luminosities, radii, masses, and ages from Gaia measurements and model assumptions, providing an independent comparison for the more information-rich LAMOST-plus-Gaia estimates.

Hubble Constant

The present expansion rate of the Universe, commonly expressed in kilometers per second per megaparsec. Early-universe analyses and local distance measurements continue to produce different values, creating the Hubble tension.

Isochrone

A theoretical curve representing stars of one age and chemical composition but different masses. Astronomers compare observed stellar temperature, luminosity, and chemistry with isochrones to estimate age, mass, and evolutionary stage.

Lambda Cold Dark Matter

The standard cosmological model, abbreviated ΛCDM. It includes ordinary matter, cold dark matter, radiation, and a cosmological constant associated with accelerated expansion. CMB-calibrated ΛCDM produces a cosmic age near 13.8 billion years.

Latent Age Distribution

The underlying distribution of true stellar ages before measurement errors blur the observations. Statistical reconstruction seeks this hidden distribution by combining the individual age likelihoods of many stars.

Metallicity

The abundance of elements heavier than hydrogen and helium in a star. Low metallicity often indicates early formation because younger stars generally formed from gas enriched by more generations of stellar nucleosynthesis and supernova explosions.

Parallax

The apparent shift in a star’s position caused by Earth’s motion around the Sun. Measuring that small angle gives distance, allowing astronomers to convert apparent brightness into intrinsic luminosity for stellar-age modeling.

Subgiant Star

A star that has begun leaving the main sequence after exhausting much of the hydrogen in its core. Its properties change quickly enough that precise distance, temperature, and chemical data can yield an informative age estimate.