Background: Studies examining risk and predictors of glucose intolerance post-GDM rely on retrospective data, or following longitudinal cohorts. Early gestational age at diagnosis, higher antepartum OGTT values and greater BMI confer greater risk of subsequent diabetes1 . However, data from women who fail to attend follow-up (around 50%2), or decline participation in longitudinal studies, are missed. It is not clear whether these women represent a group at lower/higher risk of glucose intolerance and thus whether studies over/underestimate subsequent risk.
Methods: From September 2012, women attending the GDM clinic were invited to participate in a longitudinal follow-up study. Those with pre-gestational diabetes(22), minimal English proficiency(10) and non-residents(4) were excluded. Women who returned consent were booked for OGTT at 6-12 weeks postpartum. Women who did not return consent were advised to have follow-up OGTT, and given the option of booking it at our centre. Demographic and medical details of study-consenters versus non-consenters and OGTT-attendees versus non-attendees were compared.
Results: 120 of 232 women returned consent for the longitudinal study. There were no significant differences in mean age (31.8 vs 32.5yrs), gestational age at GDM-diagnosis (24wks) or booking-in BMI (29 both groups) between those who did or did not return consent forms. Mean fasting/2hr BGL on antepartum OGTT was identical in both groups (5.1/8.9mmol/L), as was mean 3rd trimester HbA1c (5.5%). Of 112 study non-consenters, 26 completed, 26 failed to attend and 29 await postpartum OGTT. 31 are lost to follow-up. Of 120 study-consenters, 54 completed, 36 failed to attend and 30 await OGTT. Comparing 80 GTT-attendees with 62 who failed to attend, attendees were older (mean 33 vs 31yrs, p=0.01), but there was no difference in gestational age at diagnosis, antepartum OGTT or BMI. Previous history of GDM did not affect OGTT attendance (14% both groups), nor was there any difference in delivery mode, parity, insulin-use or proportion of overseas-born women (71 vs 61%, p=0.28).
Conclusions: Antepartum characteristics known to predict higher diabetes risk were not significantly different in non-study participants and non-attendees for postpartum OGTT. Non-attendance rates are high; further exploring predictors and reasons for non-attendance would enable design of strategies that improve follow-up.