SAT104 Evaluation Of Glucose Variability By Continuous Glucose Monitoring In Patients With Diabetes Mellitus And History Of Stroke

Abstract Disclosure: C. Musurakis: None. B. Poudel: None. S. Chitrakar: None. F. Qureshi: None. J.L. Gilden: None. Background: Data has now emerged that use of continuous glucose monitors (CGM) is associated with improvements in HbA1c, and patient reported outcomes. Hypothesis: In this study we assessed whether data obtained from personal CGM is associated with chronic complications of patients with diabetes mellitus (DM) and whether HbA1c was able to predict diabetic chronic complications. Methodology: A retrospective chart review of 69 DM patients [(59% T2: 8% T1: 5% of LADA) (91% with hyperlipidemia; 68% with hypertension; 59% obese )(average age= 61.47) (85% males; 15% female) (73% white: 21% black:<2% Asian; <2% Hispanic: <1% Native American) (Average HbA1c 7.74 ± 1.21) (Duration of DM >10 years =63%)(30% used insulin pumps)], from two community care-based healthcare facilities was performed. Data obtained from CGM were collected for a three-month period of time. IBM SPSS Statistics software was utilized for analysis. Results: Percentages of patients with chronic diabetic complications were as follows: 30% with urine microalbumin/cr ratio between 30-300 mg/day and 8% had >300 mg/day, 8% CKD stage III-IV, 4% ESRD, 1.4% autonomic neuropathy, 2% history of amputations, 5% history of diabetic foot infections, 34% hypoglycemia unawareness, 24% CAD, 5% had history of stroke, 7% PVD, 33% retinopathy, and 40% peripheral neuropathy. Averages ± SD of data obtained from CGM were: % time in range =52.55 ± 21.17; % high= 29.60 ± 10.89; % Very High= 15.64 ± 14.36; % very low =0.51± 1.02 and % low= 0.13 ± 0.38. The Average GMI was 7.70 % ± 0.82, with % CV 52.73 ± 18.82. A positive correlation was observed between a history of stroke and % CV, as well as GMI. AUC for %CV was 0.690 (SD 0.093, 95% CI 0.507-0.873) (p<0.05), AUC for GMI was 0.697 (SD 0.101, CI 0.499-0.895) (p=0.05). AUC for HbA1c was 0.572 (SD 0.078, CI 0.419-0.724) (NS). The % CV of 38.5 had 71% specificity and 20% sensitivity to predict stroke with a cutoff of 55.5 with 80% sensitivity and 35% specificity. The GMI of 7.65 had 80% sensitivity and 53% specificity of stroke history and a cutoff or 9.45 with 100% sensitivity and 96% specificity. HbA1c did not reach statistical significance. There were also no significant associations between HbA1c and GMI, or % CV, and other chronic complications Conclusions: Our study demonstrates that large variations of blood glucose levels, as indicated by higher than goal %CV (n<=36%), and GMI are both predictive for history of stroke. Further studies are needed to explore the pathophysiology of these associations. Furthermore, there were no associations between HbA1c and history of stroke, indicating that GMI and glucose variability, as measured by % CV, are more accurate diagnostic tools. Presentation: Saturday, June 17, 2023