Continuous glucose monitoring (CGM)
At the recruitment visit in the first trimester of pregnancy (9-13 weeks gestational age), participants will be first randomized to have either a blinded CGM sensor (sensor without a reader) (FreeStyle® LibreTM, Abbott Diabetes Care, Alameda, CA) or a non-blinded sensor (sensor with a reader) (FreeStyle Libre Pro ®, Abbott Diabetes Care, Alameda, CA) inserted on the back of either right or left upper arm on day 0, and worn up to 14 days. Glucose levels will be recorded from the interstitial fluid every 15 minutes, CGM data were downloaded from the reader for the blinded sensors, or using a software, LibreView for the non-blinded sensor users. Participants were inserted with a new sensor at the second trimester (18-23 weeks of gestational age) clinic visit. Only data from the non-blinded sensor with 70% of data captured using the sensor was used for analysis 20. The following variables were calculated from CGM readings for each participant: Mean amplitude of glycaemic excursion (MAGE), Standard deviation of blood glucose (SDBG), Mean of daily continuous 24 h blood glucose (MBG), % Coefficient of variation (CV), % of time spent in glucose target ranges and the J-Index. The % of time in target ranges were defined as: %TIR (3.5–7.8 mmol/L), %TAR (>7.8mmol/L, and %TBR (<3.5mmol/L)9. J-index, is a parameter of glucose control21, while MAGE, quantifies major swings of glycaemia and excludes minor ones was considered the gold standard for assessing intra-day glycaemic variability 22 . Extracted CGM data was used to calculate MBG, SDBG, MAGE, %CV and J-Index by an automated Software EasyGV version 9.0.R2.