Determination of time in range associated with HbA1c ≤ 6.5% in Latin American pregnant women diagnosed with type 1 diabetes mellitus using an automated insulin delivery system

Abstract

Aims: To determine the correlation between %TIR and HbA1c in pregnant women with type 1 diabetes mellitus (DM1).

Methods: Diagnostic test study in a prospective cohort of pregnant patients with DM1 using automated insulin delivery system (AID)in Colombia and Chile.

Results: Fifty-two patients were included [mean age 31.8 ± 6.2 years, pregestational HbA1c 7.2% [interquartile range (IQR), 6.5-8.2]. During follow-up, we found a better metabolic control during the second (HbA1c 6.40%, IQR 5.9,7.1) and third trimesters (HbA1c 6.25%;IQR 5.9,6.8). A weak and negative correlation between %TIR and HbA1c was found for all the gestation (Spearman's rank correlation coefficient:-0.22, p:0.0329), and in the second (r:-0.13, p: 0.38) and third trimesters (r:-0.26, p = 0.08). %TIR had poor discriminating capacity for predicting HbA1c < 6% (area under the curve [AUC], 0.59; 95% confidence interval [CI],0.46-0.72) and for predicting HbA1c < 6.5% (AUC, 0.57;95% CI,0.44-0.70). The optimal cutoff points for %TIR were > 66.1% for predicting HbA1c < 6% (65% sensitivity, 62% specificity) and %TIR > 61.1% for HbA1c < 6.5% (59% sensitivity, 54% specificity).

Conclusion: The correlation between HbA1c and %TIR during pregnancy was weak. The optimal cutoff points for identifying patients with HbA1c < 6.0% and < 6.5% were %TIR > 66.1% and > 61.1%, respectively, with moderate sensitivity and specificity.

Keywords: Continuous glucose monitoring; Glycated hemoglobin A1c; Pregnancy; Time in range; Type 1 diabetes.