A new preprint (arXiv:2604.16597) from a team including Nobel laureates Adam Riess and Brian Schmidt leverages an unprecedented low-redshift sample of 6,983 Type Ia supernova host galaxies from the TITAN DR1 catalog to test whether progenitor age differences could systematically bias cosmological distance measurements. Using spectral energy distribution fitting of ultraviolet-to-mid-infrared photometry, the researchers reconstructed star-formation histories for each galaxy, then convolved those histories with empirical delay-time distributions to infer the ages of the exploding white dwarfs' progenitor systems. This methodology enables cleaner separation of dusty star-forming versus quiescent hosts than many earlier studies achieved with smaller samples or narrower wavelength coverage.

The resulting progenitor-age distribution peaks at 2.2 Gyr with a mean of 3.5 Gyr—substantially younger than the 5–6 Gyr evolution between low- and high-redshift populations posited by models aiming to reconcile the Hubble tension through age-dependent luminosity evolution. Even when the sample is restricted to high-mass galaxies (to isolate progenitor effects from the well-documented mass step), the age gap between star-forming and quiescent systems shrinks to 3.3 Gyr. Under the strong age-evolution hypothesis, this should produce a 0.10 mag luminosity offset, yet the standardized magnitudes show no such discrepancy.

The paper therefore infers only a modest 1.5 Gyr shift in mean progenitor age across cosmic time, translating into a maximum redshift-dependent bias of ΔHR = −0.007 mag—statistically consistent with zero and already largely absorbed by standard host-mass corrections. These findings directly challenge scenarios proposed in earlier work (e.g., Rigault et al. 2013 on local star-formation rate correlations and Childress et al. 2014 on stellar population age proxies) that suggested progenitor demographics could mimic or mask cosmological signals at the level needed to erase the tension.

What much prior coverage missed is the statistical power of this large, homogeneous low-z sample and the explicit test against the mass-step degeneracy; many popular articles still treat host-galaxy correlations as an unsolved systematic capable of dissolving the 5σ Hubble discrepancy between SH0ES Cepheid+SN Ia results (≈73 km s⁻¹ Mpc⁻¹) and Planck CMB inferences (≈67 km s⁻¹ Mpc⁻¹). By synthesizing this TITAN analysis with Riess et al. (2022)'s latest SHOES H0 measurement and DESI Year-1 baryon acoustic oscillation data hinting at possible dark-energy evolution, a clearer picture emerges: standard supernova standardization techniques appear robust, pushing the community toward early-universe solutions such as early dark energy, modified gravity, or new neutrino physics rather than late-time astrophysical fixes.

As a preprint, the work has not yet undergone peer review. Limitations include its exclusive focus on z ≲ 0.15 (requiring extrapolation to match high-z SN samples), reliance on assumed universality of the delay-time distribution, and potential incompleteness in SED modeling for heavily dust-obscured systems. Nonetheless, the scale and methodological transparency of the TITAN analysis mark a substantial advance, narrowing the plausible systematic-error budget and sharpening the case that the Hubble tension is a genuine cosmological signal rather than an artifact of stellar evolution.