Is cutaneous apoprotein B a better discriminator than serum lipoproteins for atherosclerosis?