The Age of the Universe from the Oldest Stars: A New Blow to 'Early' Solutions of the Hubble Tension ✦
Based on a sample of 247,103 stars from the LAMOST DR7 survey and Gaia eDR3 parallaxes, an estimation of the age of the universe was conducted. Star ages were determined using YY isochrones (up to 20 billion years); after selecting objects with low metallicity and enhanced α-element abundances, and checking consistency with FLAME ages, the final sample consisted of 155,600 stars within 5 kpc. Reconstruction of the latent age distribution using the MCMC method yielded an age for the oldest star A★ = 13.73+0.18–0.15 billion years. This result matches the expectation from the ΛCDM model (13.8 billion years) assuming the first long-lived stars formed 0.2 billion years after the Big Bang. Such agreement challenges solutions to the 'Hubble tension' via new physics before recombination, which predict a universe age of about 12.9 ± 0.2 billion years. Uncertainties in stellar modeling cannot reconcile these values, given the low metallicity of the oldest stars and asteroseismic constraints.
Context
The age of the Universe is a fundamental quantity that defines all cosmic history. Georges Lemaître first proposed that expansion began from a super-dense state, Edwin Hubble experimentally confirmed it by discovering the redshift of galaxies, and Vera Rubin found compelling evidence for dark matter in spiral arms. Today, the standard cosmological model ΛCDM, incorporating dark matter and dark energy, predicts an age of 13.8 billion years from cosmic microwave background data. However, there is the Hubble tension—a discrepancy between the Hubble constant derived from the CMB and local measurements from supernovae. One class of solutions involves increasing the expansion rate in the early Universe, which inevitably reduces its age to about 12.9 billion years. An independent age determination from the oldest stars in our Galaxy can settle this debate. After all, if stars over 13 billion years old have managed to be born and survive to the present day, then the Universe cannot possibly be younger than that limit—the effect of cosmological time is relentless. Adding the typical time (~0.2 billion years) from the Big Bang to the formation of the first long-lived stars gives a lower bound on the age of the entire cosmos.
Methods
The scientists started with a catalog of 247,103 stars for which high-quality spectra were obtained in the LAMOST survey, and the Gaia space telescope measured precise parallaxes. The age of each star was determined by Bayesian comparison of its temperature and luminosity to evolutionary tracks (isochrones) for different abundances of helium and carbon. To filter out stars with erroneous estimates, clever population filters were applied: for example, by chemical composition—ancient stars should be iron-poor and enriched in α-elements, born in supernova explosions. Ages were also compared with independent determinations from Gaia data. The final sample included 155,600 stars within 5,000 light-years of the Sun. Then, the hidden true age distribution was reconstructed using an iterative method and Markov Chain Monte Carlo (MCMC) to account for unavoidable measurement errors.
Results
The reconstructed distribution showed a sharp drop in probability at an age of around 13.73 billion years. This value was taken as the age of the oldest star in the sample: A★ = 13.73+0.18/–0.15 billion years. Adding 0.2 billion years for the formation of the first stars gives a Universe age of roughly 13.93 billion years. This number matches with high precision the standard ΛCDM prediction: 13.8 billion years. Meanwhile, the value expected in 'early' new physics models (12.9 billion years) differs from the measured one by almost 0.8 billion years, a discrepancy of more than 3σ. Even with deliberately relaxed selection criteria, the minimum age of the oldest star was 13.31 billion years—still too high for the 'early' solutions of the Hubble tension. Analysis of a control sample of 1,556 stars with chemical signatures of the very first generations gave a similar estimate: 13.85 billion years. Thus, the result is robust to methods and criteria, and the Universe has clearly had time to grow old by our era.
Implications
This age milestone casts doubt on a whole class of explanations for the Hubble tension that invoke new physics before the recombination epoch—for example, extra relic particles or variable gravitational coupling. To simultaneously make the Universe 'young' at 12.9 billion years and avoid conflict with our data, one would have to assume a too-short interval between the Big Bang and the emergence of the first stars, which poorly matches observations of the most distant galaxies. Therefore, the resolution of the Hubble paradox likely lies in the late Universe—for instance, in local inhomogeneities ( galactic voids) or modified gravity. For the standard model ΛCDM and its components—dark matter and dark energy—the result serves as additional confirmation.
Future development
Further refinement of the oldest stars' ages will come from improved stellar evolution models, especially regarding internal mixing and helium abundance, and from new observational data. Future surveys, including the James Webb Space Telescope, will discover even older low-mass stars, while high-resolution spectroscopy will refine chemical compositions. Additionally, advances in asteroseismology will allow direct age measurements of individual 'Methuselah' stars. All this will reduce systematic uncertainties and perhaps help distinguish between competing late-time cosmological scenarios.
Impact
The results directly impact early Universe cosmology and fundamental questions about the nature of dark matter and dark energy. They are also important for galactic archaeology, which studies the formation history of the Milky Way, and for supernova theory, which determines chemical evolution.
Next steps
The researchers plan to re-analyze the data using the latest isochrones (e.g., MIST II instead of YY) and incorporate asteroseismic ages to reduce systematic uncertainties. They also intend to expand the sample with more distant stars having reliable parallaxes from future Gaia data releases.
Key open problems
The measured age of the oldest stars directly relates to the Hubble tension and constrains theories of dark energy and dark matter, since they dictate the expansion rate of the Universe throughout its history. It also sheds light on processes in the era after the Big Bang, when the first stars ignited, enriching the cosmos with helium and carbon.
🎯 The oldest star in this study is a typical subgiant of about 0.8 solar masses, which spent over 13 billion years in a calm 'middle age' (on the main sequence) and has only recently begun to swell, preparing to become a red giant. You could say we caught it just as it was about to 'retire'!
🎬 In Paul Anderson's novel 'Tau Zero,' a spaceship accelerates to near-light speed, and the crew watches as billions of years flash by outside—time nearly stops for the travelers while the Universe ages. Our work reminds us that, even without leaving the solar neighborhood, we can peer into the deep past of the cosmos, and all it takes is a careful study of stellar 'fossils'.
Key numbers
- A★: 13.73 +0.18 / -0.15 Gyr
- Age of Universe (ΛCDM): 13.8 Gyr
- Age of Universe (early solutions): 12.9 ± 0.18 Gyr
- Sample size: 155,600 stars
- Distance to stars: < 5 kpc (about 16,000 light-years)