Do gravitational wave observations in the lower mass gap favor a hierarchical triple origin?

Observations of compact objects in Galactic binaries have provided tentative evidence of a dearth of masses in the so-called lower mass gap $\sim2.2-5$ M$_\odot$. Nevertheless, two such objects have been discovered in gravitational-wave data from LIGO and Virgo. Remarkably, the estimated masses of both secondaries in the coalescences GW190814 ($m_2=2.59^{+0.08}_{-0.09}$M$_\odot$) and GW200210_092254 ($m_2=2.83^{+0.47}_{-0.42}$M$_\odot$) fall near the total mass of $\sim 2.6$ M$_\odot$ of observed Galactic binary neutron star systems. The more massive components of the two binaries also have similar masses. Here we show that a neutron star merger origin of the lighter components in GW190814 and GW200210_092254 is favored over $M^{-2.3}$ (Bayes factor $\mathcal{B}\sim 5$) and uniform ($\mathcal{B}\sim 14$) mass distributions in the lower mass gap. We also examine the statistical significance of the similarity between the heavier component masses of GW190814 and GW200210_092254, and find that a model in which the mass of GW200210_092254 is drawn from the mass posterior of GW190814 is preferred ($\mathcal{B}\sim 18$) to a model in which its mass is drawn from the overall mass distribution of black holes detected in gravitational wave events. This hints at a common origin of the primary masses, as well as the secondary masses, in GW190814 and GW200210_092254.