Do lowest-order approximations adequately describe nuclear matter?