A changing dark-energy model reshapes neutrino-mass estimates—but doesn’t win every test

Cosmology’s standard model, known as ΛCDM, describes dark energy as a constant contribution to the Universe’s expansion. Recent analyses combining measurements from the Dark Energy Spectroscopic Instrument (DESI) with other observations have prompted interest in alternatives where dark energy changes over time. Neutrino mass adds another puzzle: some cosmological fits within ΛCDM have placed the preferred value near zero or even in an unphysical negative range, despite laboratory measurements of neutrino oscillations implying a nonzero minimum mass sum.

Mario A. Rodríguez-Meza, Jorge L. Cervantes-Cota and Tonatiuh Matos investigate whether a particular dark-energy extension can address both questions. Their paper is an arXiv preprint, so its findings should not be taken as independently confirmed.

A time-dependent vacuum-like component

The authors call their proposal conformal holographic dark energy, or CHDE. In the model, dark-energy density follows a power law in conformal time, a time coordinate used in cosmology. The power-law exponent, n, is the model’s extra parameter: when n is zero, CHDE reduces to ΛCDM.

Although CHDE retains an equation of state of −1—like a cosmological constant—it allows energy to pass between dark energy and dark matter. In the version studied, the preferred negative values of n mean that dark energy declines over time and acts as a source of dark matter. This is an implication of the model, not a direct observation of energy moving between the components.

The researchers implemented the model in the cosmology code CLASS and used Cobaya to sample possible parameter values with Markov chain Monte Carlo. They compared predictions with baryon acoustic oscillation (BAO) distances from DESI’s second data release, cosmic microwave background measurements from Planck, gravitational lensing of the microwave background measured by the Atacama Cosmology Telescope, and Type Ia supernova distance data. Different analyses used different combinations of these observations. The authors considered both flat universes and models that allow spatial curvature, and in some combinations allowed the summed neutrino mass to vary.

A preference in parameter fits, but not a clear model-selection win

Across several combinations, the inferred exponent is around −0.25 to −0.32, with quoted uncertainties generally under 0.1. The paper describes the ΛCDM parameter choice, n = 0, as disfavored at roughly 4.4–4.5 standard deviations in particular analyses. That figure measures the separation of the specified ΛCDM point from the fitted CHDE parameter distribution under the authors’ calculation. It is not a general 4.4–4.5-sigma proof that CHDE is the correct explanation.

The paper’s separate model-comparison tests do not consistently favor CHDE. For the version allowing curvature, one information criterion mildly favors CHDE while another favors ΛCDM. For a flat universe, fit quality is nearly the same, and both reported information criteria favor ΛCDM. These tests account differently for the extra flexibility in a model. Their mixed results matter: a parameter value can differ from zero without the extended model being clearly preferred overall.

In the CHDE analyses, the neutrino-mass posterior distributions peak at positive values. But compatibility with terrestrial lower bounds depends on the dataset combination and assumptions. For example, the flat analysis combining BAO, Planck, ACT and Union3 reports a mass sum of 0.018 eV, with a 95% upper endpoint of 0.051 eV—still below the cited 0.059 eV lower bound for the normal mass ordering. A positive peak therefore does not mean every reported interval reaches the terrestrial bounds. The authors note that BAO, Planck and other datasets do not all favor the same matter-density ranges; their limited overlap affects the combined mass constraint. They also caution that a different prescription for dark-energy perturbations could shift the inferred masses.

What remains uncertain

CHDE is a phenomenological proposal: its power law is motivated by earlier horizon-based models, but the paper does not derive that particular form from a fundamental theory. The analysis also adopts a specific prescription for perturbations and energy transfer. Those choices affect the model’s predictions and can influence inferred parameters.

The authors say the modified dark-matter perturbations could tend to suppress the growth of matter clumps, but leave a dedicated analysis of growth observables for future work. That possible effect should not be mistaken for a measured result of this study.

The narrow conclusion is that, within the datasets, assumptions and statistical methods tested, CHDE fits favor a negative exponent and produce neutrino-mass distributions peaking at positive values. The reported mass constraints are not uniformly compatible with terrestrial lower bounds, and the model-comparison statistics provide no consistent preference for CHDE over ΛCDM. Further analyses will be needed to assess how robust these results are.