Does DKA at Diagnosis Predict Better Glycemic Control in Patients with Type I Diabetes Mellitus?

Presentation Number: SAT 586
Date of Presentation: April 1st, 2017

Marina Epelman*1, Lily Mohtadi2, Julianne Pereira3 and Siham D Accacha1
1Winthrop University Hospital, Mineola, NY, 2Cornell University, 3Northwell Health Hospital


Background: Incidence of type 1 diabetes mellitus (T1DM) in children has been increasing worldwide by 3% to 5% per year. Approximately 29.4% of newly diagnosed patients present with ketoacidosis. Good glycemic control in first years after diagnosis has been proven to preserve the residual secretion of insulin and associated with lower risk of hypoglycemia. Previous studies have suggested that poor glycemic control is associated with older age, female sex, disease duration, poor patient provider relationship. Predictor of good control are: basal bolus regimen, young age at onset, higher socioeconomic status, family support. However, little is known about association of DKA at diagnosis and future diabetes control.

Objectives/Hypothesis: We hypothesize that patient who had DKA at diagnosis have better glycemic control compared to patients with incidentally diagnosed T1DM due to better adherence to treatment.

Methods: We reviewed 134 charts of patients 0-18 years old who were diagnosed with T1DM between July 1st, 2009 and June 30th, 2012 at Winthrop University Hospital. Levels of HbA1, average blood glucose monitoring (BGM), lipid panel results were collected from diagnosis to a time 2 years following diagnosis of T1DM. Statistical analyses were done in SAS 9.4®.

Results: From 134 patients, enrolled in the study, 37 (27.6%) patients had DKA at diagnosis.

There was no significant difference in age, gender, and ethnicity among the two groups. Baseline HbA1c was significantly higher in DKA group compared to non-DKA (12.0 ± 2.5 and 11.0 ± 2.1 respectively, p=0.04). In mixed effect model there was no significant difference in improvement of HbA1C over time among the two groups. Furthermore, there was no difference in HbA1C between two groups at the end of 24 month follow up period in unadjusted and adjusted ANCOVA analysis.

Conclusions: DKA at diagnosis of T1DM was not predictive of glycemic control at 2 year follow up.


Nothing to Disclose: ME, LM, JP, SDA