Graduation Year


Document Type




Degree Name

Master of Science (M.S.)

Degree Granting Department

Public Health

Major Professor

Janice Zgibor, Ph.D.

Committee Member

Amy Alman, Ph.D.

Committee Member

Skai Schwartz, Ph.D.


Diabetes, HbA1c, Hyperglycemia, Hypoglycemia


Introduction: Both hypoglycemia (low blood glucose) and hyperglycemia (high blood glucose) are common among individuals with type 1 diabetes and are associated with severe medical complications, therefore it is essential that health care providers are able to accurately measure glycemic control. Measures derived from continuous glucose monitoring (CGM) may provide more accurate measurements of glycemia than the commonly used HbA1c blood test.

Methods: Data from the Juvenile Diabetes Research Foundation (JDRF) clinical trial to assess the efficacy of continuous glucose monitoring was used to estimate the ability of CGM composite scores to predict time in range, time in hyperglycemia and time in hypoglycemia. Spearman partial correlation coefficients were calculated between composite measures and thresholds of glycemia.

Results: HbA1c showed week correlations with time below 54 mg/dL (R = -0.05) and time below 70 mg/dL (R = -0.21). HbA1c was moderately correlated with time in range (R = -0.62), however, most other CGM derived metrics had stronger correlations. HbA1c was also moderately correlated with time above 180 mg/dL (R = 0.68) and time above 250 mg/dL (R = 0.64), however several CGM derived composite scores including the J-index, GRADE, the Q-score, GMI, CGP, and PGS all had stronger correlations with time in hyperglycemia (Table 3).

Conclusions: HbA1c is a poor predictor of time in hypoglycemia and moderately correlated with time in hyperglycemia. Several composite metrics had stronger correlations with both hypo and hyperglycemia than HbA1c and were better predictors of meeting clinical targets.

Included in

Epidemiology Commons