Comment on the paper: ``Feynman-Schwinger representation approach to nonperturbative physics'' by C. Savkli et al. [Phys. Rev. C 60, 055210 (1999), hep-ph/9906211]

We point out that the scalar Wick-Cutkosky model (with ${\chi }^2\varphi$ interaction) used in this study has been known for a long time to be unstable. However, the numerical results presented in this paper do not show any sign of this instability which casts some doubt on their reliability. We compare with the worldline variational approach.