Slow Massive-Star Formation and the O-Star ZAMS Deficit

The scarcity of optically visible O stars near the theoretical zero-age main sequence (ZAMS) has puzzled astronomers for over four decades. We argue that it reflects prolonged accretion. If massive stars assemble over Myr timescales, as the inertial-inflow model predicts, core-hydrogen burning begins before accretion ends, and the stars become optically visible at the end of accretion on a birthline displaced from the standard ZAMS. We test this interpretation by comparing the hot-side envelope of Galactic O stars in the Kiel diagram with accreting stellar-evolution models, adopting a growth law $t_{\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{m}}={\tau }_0(m_{\mathrm{f}}\mathrm{/}60M_{\odot }{)}^{\alpha }$ with formation time $t_{\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{m}}$ and final mass $m_{\mathrm{f}}$. The resulting birthlines reproduce both the displacement and the high-mass turnover of the observed envelope, with best-fitting parameters $(\alpha ,{\tau }_0)=(0.52,4.3\mathrm{M}\mathrm{y}\mathrm{r})$, consistent with independent constraints from the luminosity functions of compact H II regions and OB stars. The two constraints have orthogonal degeneracy directions, so combining them narrows the allowed parameter space, yielding projected 90% confidence intervals $0.43<\alpha <0.60$ and $3.9<{\tau }_0\mathrm{/}\mathrm{M}\mathrm{y}\mathrm{r}<5.0$. The inferred growth law agrees with cloud-scale simulations and the inertial-inflow model, and implies that mass assembly and main-sequence evolution overlap substantially in massive stars, with broader implications for stellar evolution and feedback.