The model’s Dark Star scenario becomes especially relevant when the stars grow to around, or beyond, $10^6$ solar masses. Collapse at that scale can provide a heavier starting point for black-hole growth than conventional light-seed scenarios. If enough of those remnants form and later participate in mergers, they can substantially raise the predicted PTA background.
The result is therefore conditional. It does not establish that supermassive Dark Stars existed, nor does it show that the PTA signal uniquely comes from their remnants. It shows that a sufficiently numerous population of very massive Dark Stars could leave descendants capable of contributing strongly to the signal.
Ghodla and Ilie compared Dark-Star remnants with direct-collapse black holes (DCBHs), another proposed route to heavy seeds in the early Universe. In the model summarized by the study, Dark-Star remnants with a comoving number density of about $10^{-3},\mathrm{Mpc}^{-3}$ can provide a significant—and potentially dominant—share of the PTA background. The DCBH channel, using the much lower seed densities adopted for that comparison, contributes only marginally or remains subdominant.
The difference reflects the formation requirements built into the two scenarios. DCBHs generally require unusual environments in which pristine gas collapses without fragmenting, often under intense ultraviolet radiation from nearby star formation. The Dark-Star setup can allow a more numerous population if suitable isolated minihalos are common and if the objects can keep accreting to supermassive scales.
The comparison does not rule out DCBHs. It indicates that, under the study’s assumptions, the abundance of seeds matters as much as their individual masses: a more numerous Dark-Star-remnant population can generate a larger late-time binary background than a rarer heavy-seed channel.
The same calculation places an upper constraint on the early-seed population. In the framework used by the authors, densities of roughly $10^{-2}$–$10^{-1},\mathrm{Mpc}^{-3}$ would produce too much gravitational-wave energy compared with the observed PTA background. The data therefore disfavor seed populations in that range under the assumption that the signal is primarily generated by supermassive black-hole binaries.
This is a population-level constraint, not a direct census of Dark Stars. Its strength depends on assumptions about seed formation, accretion, halo-merger histories, binary pairing and evolution, and the interpretation of the PTA signal itself. PTAs are detecting—or constraining—a gravitational-wave background; they are not resolving individual primordial stars.
The broader significance is that gravitational-wave astronomy can probe objects that may have disappeared long ago. Electromagnetic searches could look for unusually bright or massive objects at high redshift, while PTAs test the merger history of the black holes those objects might have produced.
If future PTA measurements refine the background’s amplitude and spectrum, and if the supermassive-binary interpretation remains favored, those results could help distinguish among Dark-Star remnants, DCBHs and other early-seed channels. The study’s central contribution is thus a bridge between cosmic-dawn physics and present-day nanohertz observations: the background may preserve information about how the first massive black holes were seeded, even if the original Dark Stars themselves are no longer visible.