The IMF describes the distribution of stellar masses at birth—how many low-mass stars (like red dwarfs) form compared to high-mass stars (like blue giants). The key feature the Steinhardt team measured is the IMF's "break mass": the mass where the power-law slope of the distribution changes .
Using Gaia's ultra-precise measurements of stars in 110 high-quality open clusters, the researchers found that the break mass varies considerably between clusters, and that variation tracks with the age of the cluster . This is direct evidence that the IMF is not a fixed law of nature but is instead shaped by local star-forming conditions. The study builds on earlier evidence: Kirkpatrick et al. (2023) showed using Gaia data that the log-normal form of the IMF fails below about 0.3 solar masses
, and a 2025 census of approximately 3,600 stars within 20 parsecs found a different low-mass IMF shape than the canonical form
. A 2026 study of the same 110 open clusters further demonstrated that the ratio of high-mass to low-mass stars changes with cluster environment
.
The leading physical explanation for IMF variation involves the Jeans mass—the minimum mass a collapsing gas fragment must have to overcome internal pressure and form a star. The Jeans mass depends on the gas temperature and sound speed: in warmer, more turbulent gas, the Jeans mass rises, pushing the IMF toward top-heaviness (more high-mass stars relative to low-mass ones). In cold, quiescent gas, the Jeans mass is lower, favoring a bottom-heavy IMF with more low-mass stars .
This means that the environment where stars are born—its temperature, density, metallicity, and turbulence—directly controls what kind of stars form. The Milky Way's IMF, long treated as the universal standard, is just one local example among many possible IMF shapes.
This discovery lands at a perfect storm moment in astronomy. JWST has been revealing galaxies at redshifts z > 7 that appear far too massive and too numerous to fit within the standard ΛCDM cosmological model . These galaxies challenge our understanding of how quickly structure can assemble in the early universe—unless we've been using the wrong IMF to weigh them.
Here’s the critical mechanism: if the IMF in the early universe was top-heavy (more high-mass stars per unit of star formation), then galaxies would produce far more ultraviolet light per unit of stellar mass than a Milky Way-like IMF. Current spectral energy distribution (SED) fitting codes assume a universal IMF, so when they see lots of UV light, they infer lots of stellar mass. But if the IMF was top-heavy, that light is coming from fewer, more luminous stars, not from more total stellar mass .
The numbers are dramatic. Woodrum et al. (2024), using JADES NIRCam photometry, showed that models with a top-heavy IMF reduce inferred stellar masses by up to a factor of three compared to the standard assumption, and that these models remain fully consistent with the observed SEDs . The COSMOS2020 catalog analysis found that galaxies exhibit a continuum of IMF shapes, mostly bottom-lighter than the Milky Way, with stellar masses lower by factors of 1.6–3.5 and SFRs lower by factors of 2.5–70 compared to traditional techniques
.
Cosmological simulations are catching up fast. The COLIBRE galaxy formation model, run in a 100 comoving Mpc cosmological simulation incorporating a density-dependent IMF, naturally reproduces the JWST-observed high-redshift galaxy counts without requiring exotic physics . The GAEA model similarly shows that a variable IMF framework can explain the puzzling space densities of bright galaxies at z > 7
.
Stellar mass corrections: Early galaxies are likely less massive than we think—possibly by factors of 3–10—bringing them into better agreement with ΛCDM predictions and reducing the "tension" that has generated headlines since JWST's first deep fields .
Star formation rate overestimates: Because UV luminosity calibrations assume a fixed ratio of UV light to star formation, a top-heavy IMF means current SFR estimates for high-redshift galaxies are likely too high. The true star formation efficiency may need to be revised upward if fewer total stars formed .
Systematic uncertainty in galaxy evolution: Every mass-dependent galaxy property—stellar mass functions, star formation main sequence, mass-metallicity relation—carries an IMF-driven systematic error whose direction and magnitude can now, for the first time, be directly measured and corrected using Gaia's empirical results .
The end of the universal IMF assumption does not break astronomy—it makes it more precise. We now know the IMF is a function of environment, and Gaia has given us the data to start measuring that function empirically. The next steps are to calibrate IMF variation across different environments (temperature, metallicity, redshift) and build that variation into SED fitting codes for JWST and next-generation observatories like the Nancy Grace Roman Space Telescope.
The message for cosmology is clear: the "too massive, too early" problem may not require new physics—it may simply require us to stop applying a Milky Way rulebook to the early universe. Gaia has given us the evidence; it's time to rewrite the rules.