This short communication analyses a boundedness property of the inverse of a Jacobian matrix that arises in regularized primal-dual interior-point methods for linear and nonlinear programming. This result should be a useful tool for the convergence analysis of these kinds of methods.