Ketoacidosis at onset of type 1 diabetes is a predictor of long-term glycemic control

Background: Few studies have evaluated the impact of diabetic ketoacidosis (DKA) at diabetes onset on long-term glycemic control in patients with type 1 diabetes (T1D). Objective: We aimed to determine any differences in long-term glycemic control between children/adolescents with T1D presenting with DKA at diabetes onset and those without. Methods: This retrospective study comprised 335 patients diagnosed with T1D from September 2007 to December 2012, among which 132 (39.4%) presented with DKA. Variables compared between patients with DKA at onset and those without: yearly hemoglobin A1c (HbA1c) levels, daily insulin dose, yearly rates of severe hypoglycemia and DKA, percent of patients achieving target HbA1c levels. Results: After the first year of diabetes, the mean daily insulin dose and HbA1c level were significantly higher in the group with DKA at onset (0.74 ± 0.26 vs 0.69 ± 0.27 units/kg/d, P =.049, and 7.85 ± 1.13% vs 7.49 ± 0.94%, P =.01, respectively), despite similarity of therapy (multiple daily injections or continuous subcutaneous insulin infusion), with a similar but not statistically significant trend subsequently. Mean HbA1c since onset was significantly higher in the DKA group (8.08 ± 0.95% vs 7.86 ± 0.95%, P =.025). A significantly higher percentage of patients in the group without DKA at onset achieved a mean level of HbA1c since onset within glycemic targets (32% vs 20.5%, P =.02). In the DKA group, the frequency of subsequent DKA episodes per diabetes years was significantly higher (P =.042). Conclusions: DKA at diagnosis was associated with less favorable long-term glycemic control as assessed by HbA1c and the rate of DKA episodes. T1D patients presenting with DKA may therefore need stricter treatment and tight follow-up.