Comments on variational approximations within the hydrodynamic model