1College of Medical and Dental Sciences; [email protected]; 2University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK; 3RCSI & UCD Malaysia Campus, Penang, Malaysia; 4University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK & Institute of Metabolism and Systems Research, University of Birmingham, UK; 5College of Medical and Dental Sciences, University of Birmingham, UK; 6University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK; 7University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK & Institute of Immunology and Immunotherapy, University of Birmingham, UK
Objectives: Limited evidence exists studying the impact of ethnicity on diabetic ketoacidosis (DKA). We aimed to study the impact of ethnicity on presentation, management and outcome of DKA in type 1 (T1DM) and type 2 (T2DM) diabetes.
Methods: All DKA episodes from April 2014 to September 2020 in a UK tertiary care centre were identified. Data were collected on diabetes type, demographics, biochemical and clinical features on admission, and DKA management. StataSE16 was used for analysis. As data were skewed, median and interquartile range (IQR) are presented with Wilcoxon sum rank test used to compare groups.
Results: 583 (75.9%) T1DM and 185 (24.1%) T2DM episodes were included. Non-white ethnic groups were significantly overrepresented in T2DM (proportion of Whites in T1DM 80.3% vs T2DM 73.5%; P=0.030). In people of White ethnicity, those with T1DM presented more hyperglycaemic than those with T2DM (T1DM 28.20 [IQR 20.45-34.25] vs T2DM 23.60 [15.61-29.60]; P=0.01). Conversely, in those of a Black ethnicity, patients with T2DM diabetes presented more hyperglycaemic (T1DM 22.90 [18.30-37.30] vs T2DM 38.80 [31.10-39.60]; P=0.03). People of Asian ethnicity had higher lactate levels in T2DM (T1DM 2.54 [1.91-3.80] vs T2DM 4.20 [3.40-6.17]; P=0.01). There were no differences in DKA duration, though length of hospitalisation was greater in those with T2DM in all ethnicities apart from Blacks. When data was disaggregated by diabetes type, statistically significant differences were seen between ethnicities in urea in T1DM and bicarbonate, lactate and serum osmolality in T2DM, but not DKA duration or length of hospitalisation.
Conclusions: Non-white ethnicities are more likely to present with DKA in T2DM. In non-White ethnicities, DKA may present with more severe biochemical abnormalities in T2DM than in T1DM. While there are differences in presentation and severity of DKA between ethnicities in T1DM and T2DM, management and outcome of DKA were unaffected.