Reproducibility of in-home CFRD screening using continuous glucose monitoring and mixed meal tolerance test

Cystic fibrosis related diabetes (CFRD) is associated with insulin-remediable pulmonary decline, so early detection is critical. Continuous glucose monitors (CGM) have shown promise in screening but are not recommended by clinical practice guidelines. Little is known about the reproducibility of CGM results for a given patient.

Twenty non-insulin treated adults and adolescents with CF placed an in-home CGM and wore it for two 14-day periods. Participants underwent a mixed meal tolerance test (MMTT) on day 5 of each 14-day period. Glycemic data from CGM 1 and CGM 2 were compared regarding published thresholds to define abnormality: percent time >140 mg/dL of ≥4.5%, percent time >140 mg/dL of >17.5%, and percent time >180 mg/dL of >3.4%. Results of the repeat MMTT were compared for peak glucose and 2-hour glucose thresholds: >140 mg/dL, >180 mg/dL, and >200 mg/dL.

For percent time >140 mg/dL of ≥ 4.5%, five of 20 subjects had conflicting results between CGM 1 and CGM 2. For percent time >140 mg/dL of >17.5% and >180 mg/dL of >3.4%, only one of 20 subjects had conflicting results between CGM 1 and CGM 2. On the MMTT, few participants had a 2-hour glucose >140 mg/dL. Peak glucose >140 mg/dL, 180 mg/dL, and 200 mg/dL were more common, with 10-37% of participants demonstrating disagreement between CGM 1 and CGM 2.

Repeated in-home CGM acquisitions show reasonable reproducibility regarding the more stringent thresholds for time >140 mg/dL and >180 mg/dL. More data is needed to determine thresholds for abnormal mixed meal tolerance tests in CFRD screening.