Standardization of methods in the epidemiology of diabetes mellitus.

While considerable advances have been made in devising standardized methods for epidemiologic studies, it is only recently the standardized methods for use in diabetes epidemiology have been seriously considered. It is hoped that the present paper will provoke further attention to these matters and result in further discussions of areas where it appears possible to reach a consensus. Research into the performance of certain of the methods mentioned is urgently needed, and it is hoped that epidemiologists interested in diabetes will take up this challenge to prove or disapprove that some of the recommendations and suggestions made in this paper are useful and worthwhile.