This original research from Jothydev’s Diabetes Research Centre was presented at Advanced Technologies and Treatments for Diabetes (ATTD) on February 27-Mar 2,2013 at Paris, France.

Background and aims: Glycemic Variability (GV) in addition to HbA1c and serum glucose levels is emerging as an independent but powerful risk factor contributing to endothelial damage and vascular complications. Most of the published studies on Continuous Glucose Monitoring (CGM) and measurements on GV are in type 1 diabetes. Considering the huge number of subjects with type 2 diabetes and on different therapies, it is worthwhile to look at SD, MAGE and MODD. With advancing stages in
type 2 diabetes(T2DM), when subjects are on multiple daily injections,

GV could be a significant factor to be considered with an intention to prevent vascular complications. This is supposed to be the first study from India to analyze the GV from CGM data.

Material and methods: In this exploratory study, we recruited 63 T2DM patients on treatment in our clinic for continuous glucose monitoring. We calculated three measures of glycemic variability from the CGMdata that yielded at least 48 hours of continuous data with 288 measurements per day. The measures calculated were SD, MAGE and MODD. Data of 59 subjects was analyzed, (49 male, 10 female), with mean age 54 years (SD 11.5, range 29-79), mean duration of diabetes 11±5.7 years, and mean HbA1c 7.8±1.3%. Seven patients were on OHA only; insulin regimens for others were basal (n=7), basal-bolus (17), biphasic (12) and CSII (16), with mean total daily dose of insulin (TDD) 28±21 U. Results: Analysis showed mean SD 38.5±17, mean MAGE 76±32 mg/dL, and mean MODD 36±17 mg/dL. There was a highly statistically significant correlation among the three measures of variability (MAGE: MODD 71%; MAGE:SD 95%, MODD:SD 79%). Bivariate analysis showed increasing SD with increasing TDD (p<0.000) and a positive trend with increasing age (p=0.07). Differences in SD by treatment group and sex were not statistically significant despite large differences (OHA 28.6 vs basal-bolus 42.1; females 49 vs males 36). Sulfonylurea use was associated with higher SD (mean SD 42.9 vs 32.3; p=0.02). Conclusion: This pilot study yields the first data towards setting benchmark levels of glycemic variability in South Indian population. Compared to biphasic and basal-bolus regimens, insulin pump therapy was observed to have the least GV. Larger studies are needed to study the trends of association of GV with female sex and level of differences in GV with different insulin regimens. The study shows higher total daily dose of insulin are associated with increasing SD, independent of duration of diabetes. Association of sulfonylureas with glycemic variability is also an important finding that needs to be confirmed, as there are implications for prevention of long-term complications.