We further excluded women who reported a physician’s diagnosis of diabetes (n = 22) or borderline diabetes treated by oral hypoglycemics (n =2), women reporting previous borderline GDM (n = 17), ฃ(n = 168; 18 did neither recall, 147 completed only 1 recall, and 3 had a second recall that was deemed unreliable by interviewers), resulting in an analytic cohort of 817 women.