DESI Data Squeeze Dark Energy’s Secret Handshake With Dark Matter to the Brink of Detection

For nearly three decades, cosmologists have lived with an uncomfortable truth: the standard model of the universe works almost too well. The Lambda cold dark matter model, in which a mysterious cosmological constant drives cosmic acceleration while cold dark matter binds galaxies together, fits an extraordinary range of observations. Yet the model leaves deep theoretical scars, including the so-called coincidence problem of why dark matter and dark energy densities are comparable today, and the lingering Hubble tension over the universe’s expansion rate. Now, a new analysis published in The European Physical Journal C has put one of the most popular alternatives to Lambda-CDM through the most demanding observational gauntlet yet, and the verdict is striking: dark matter and dark energy appear to keep their distance, at least within the limits of today’s data.

The study, led by Fan Yang of Hunan University of Science and Technology together with colleagues in China, examines interacting dark energy, a class of models in which the dark sector is not passive but conversational. Instead of evolving independently, dark matter and dark energy exchange energy through a non-gravitational coupling, a handoff that could naturally produce the effective dynamical dark energy behavior hinted at by recent surveys. The researchers considered two versions of this idea: an interacting variant of wCDM, where dark energy’s pressure-to-density ratio, its equation of state w, is a constant, and an interacting version of the Chevallier-Polarski-Linder model, in which w evolves over cosmic time according to two parameters, w0 and wa.

The technical engine of the analysis is the growth of matter density perturbations, the subtle wrinkles in the early universe that gravity amplified into galaxies and galaxy clusters. Cosmologists summarize this growth with the growth rate f, defined as the logarithmic derivative of the matter density contrast with respect to the scale factor. A widely used shorthand links f to the matter density fraction via a power law, f approximately equal to Omega_m raised to the growth index gamma. In standard Lambda-CDM governed by General Relativity, gamma sits at roughly 0.545. The beauty of this number is its diagnostic power: modified gravity theories predict different values, with the DGP braneworld model yielding about 0.6875 and viable f(R) gravity models clustering near 0.40 to 0.43.

Here is where the physics becomes genuinely subtle. The team derived a second-order approximation for the growth index in the interacting wCDM scenario and uncovered a clean theoretical result: the dark-sector coupling alpha induces a nearly universal correction to the growth index of about 1.1 times alpha, in both the wCDM and CPL frameworks. That means a positive coupling, in which energy flows from dark energy into dark matter, pushes the growth index upward toward modified gravity territory, while a negative coupling drags it down toward the f(R) range. Theoretical work dating back to studies of interacting quintessence has shown that such a coupling can even be engineered to reproduce both the expansion history and the growth history of DGP gravity exactly, rendering the two fundamentally different explanations observationally indistinguishable from expansion and growth data alone. If the coupling is free to vary, the growth index loses its diagnostic punch.

To determine whether observations actually permit this mimicry, the researchers confronted both models with an arsenal of modern cosmological data. The background expansion probes included the Pantheon+ compilation of 1701 Type Ia supernova light curves spanning redshifts from 0.001 to 2.26, thirty-two Hubble parameter measurements from cosmic chronometers, compressed distance priors from the Planck 2018 cosmic microwave background release, and two alternative baryon acoustic oscillation datasets, one drawn from all four generations of the Sloan Digital Sky Survey and one from the second data release of the Dark Energy Spectroscopic Instrument. Crucially, they folded in redshift-space distortion measurements of the quantity f sigma 8, which combines the growth rate with the amplitude of matter fluctuations and is immune to the galaxy bias that plagues raw growth observations.

On the theoretical side, the team was careful to handle a known pitfall. In the interacting case, the simple constant-gamma approximation breaks down: at low redshifts the discrepancy between the parameterization and the true growth rate reaches roughly 12 percent for a coupling of 0.02, comparable to the typical 15 percent observational error on f sigma 8. Their improved second-order formula for gamma(z), which explicitly includes the coupling alpha, cuts that maximum error to about 5 percent, a roughly 60 percent improvement. They validated this analytical shortcut against exact numerical integration of the modified growth equation within Markov Chain Monte Carlo analyses, finding robust agreement in both central values and uncertainties, and gaining substantial computational efficiency in the process. The analysis also explicitly included the radiation component in the background equations to guarantee accuracy at high redshifts, which matters for the CMB distance priors.

The results are a triumph of precision. Using the DESI DR2 baryon acoustic oscillation data combined with the other background probes and redshift-space distortions, the interacting wCDM model yields a coupling constant of alpha = 0.0230 plus or minus 0.0076, consistent with zero at approximately the three sigma confidence level, and an equation of state w = -0.948 plus or minus 0.025, consistent with the cosmological constant value of -1 at two sigma. The dynamical CPL version tells a similar story: alpha = 0.0173 plus or minus 0.0095, with w0 = -0.888 plus or minus 0.059 and wa = -0.28 plus or minus 0.25, all consistent with their Lambda-CDM values within two sigma. In short, there is no definitive statistical evidence that dark matter and dark energy are talking to each other, though the DESI-based constraint on the coupling is roughly 85 percent tighter than comparable earlier analyses.

The implications for the degeneracy with modified gravity are decisive. Translating the observational bounds on alpha into the growth index, the DESI-based wCDM analysis restricts gamma to the interval from about 0.5469 to 0.5974 at three sigma, while the CPL model yields roughly 0.53 to 0.59. Taken together, both models confine the growth index to a common window of approximately 0.53 to 0.60. That window is far too narrow to reach either the DGP prediction near 0.69 or the f(R) predictions near 0.40 to 0.43. The region of parameter space where interacting dark energy could masquerade as modified gravity is strongly disfavored by current data, which means the growth index regains its power as a diagnostic despite the theoretical degeneracy.

The analysis also revealed telling correlations among the dark sector parameters. In the interacting wCDM model, the coupling alpha and the equation of state w are positively correlated: a stronger energy transfer to dark matter prefers a less negative equation of state to preserve the observed expansion history. In the CPL model the pattern grows more intricate, with a negative correlation between w0 and alpha but a positive one between wa and alpha. Interestingly, allowing the equation of state to vary in time partially absorbs the signatures that would otherwise be attributed to the interaction, shifting the best-fit coupling closer to zero and slightly widening its uncertainty. This warns that fixing any of these parameters to their Lambda-CDM values could bias the others and mask the true underlying mechanism, so future analyses must constrain the dark sector jointly.

Looking ahead, the study sharpens the interpretive framework for the coming decade of structure surveys. The Dark Energy Spectroscopic Instrument’s dedicated redshift-space distortion analysis, the Euclid space telescope, and the Square Kilometre Array will measure the growth of cosmic structure with unprecedented precision. The message from this work is clear: if those experiments detect a statistically significant deviation of the growth index from the Lambda-CDM value of about 0.55, the culprit will most likely be a genuine modification of gravity on cosmological scales rather than a clandestine exchange between dark matter and dark energy, whose permitted shifts now appear too small to account for such a signal. Meanwhile, with recent DESI and Dark Energy Survey analyses pointing toward mildly evolving dark energy, the question of what actually drives cosmic acceleration remains wide open, and the universe’s dark components will keep their secrets a little longer.

Subject of Research: Observational constraints on interacting dark energy models and the growth index of matter density perturbations

Article Title: Impact of interacting dark energy on the growth of matter density perturbations: observational constraints from DESI and multi-probe data

Article References: Yang, F., Zhai, R., Fu, X., Xu, B., Liu, K., Ding, C., & Huang, Y. (2026). Impact of interacting dark energy on the growth of matter density perturbations: observational constraints from DESI and multi-probe data. The European Physical Journal C, 86(9), Article 1115. https://doi.org/10.1140/epjc/s10052-026-16366-9

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16366-9

Keywords: dark energy, dark matter, DESI, growth index, modified gravity, baryon acoustic oscillations, redshift-space distortions, Pantheon+, cosmic microwave background, Lambda-CDM, CPL parameterization, cosmology