Admissions for diabetic ketoacidosis in ethnic minority groups in a city hospital Ebenezer Nyenwe a, 4 , Raghu Loganathan a , Steve Blum b , Donald Ezuteh a David Erani c , Marcia Palace c , Chukwuma Ogugua a a Department of Medicine, Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, Bronx, NY 10457, USA b Department of Clinical Epidemiology, Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, Bronx, NY 10457, USA c Division of Endocrinology, Diabetes and Metabolism, Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, Bronx, NY 10457, USA Received 3 February 2006; accepted 18 September 2006 Abstract Hospitalization for diabetic ketoacidosis (DKA) is increasing, perhaps due to the rising incidence of DKA in patients with type 2 diabetes mellitus (T2DM). Ethnic minority groups are at increased risk for T2DM. This study aimed at elucidating the characteristics of patients with ketosis-prone diabetes in a predominantly ethnic minority population. We performed a retrospective analysis of adults admitted with DKA at the Bronx Lebanon Hospital Center, Bronx, NY over 3 years. The patients were divided into cohorts based on type of diabetes and ethnicity. The cohorts were described and compared using statistical methods. We recorded 219 cases of DKA in 168 patients, 97% of whom were African American or Hispanic. Fifty-three (32%) patients had T2DM. New-onset diabetes, which was more common in T2DM ( P b .0001), and African Americans ( P = .008), occurred in 42 patients (25%). Readmission with DKA was more common in the Hispanic patients with type 1 diabetes mellitus (T1DM) ( P = .0001). Type 2 diabetes mellitus was more prevalent in the African Americans ( P = .04). Patients with T1DM had more severe acidosis than patients with T2DM (lower pH and bicarbonate and larger anion gap; P = .03, .02, and .005, respectively). Creatinine level was higher in patients with T2DM ( P = .04) who were also less likely to have identifiable precipitating causes ( P = .02). Hemoglobin A 1c level was higher in patients with new-onset diabetes ( P b .05), but did not differ between those with T1DM and T2DM. Mortality, which was 2%, occurred only in the African Americans with T2DM. We conclude that DKA is an important mode of initial presentation of T2DM, with new-onset T2DM accounting for about 60% of all new cases of DKA. African American patients with T2DM, in comparison with the Hispanic patients, are more susceptible to developing DKA. Diabetic ketoacidosis could occur in T2DM without any identifiable precipitant. The rising incidence of DKA may be attributable to its increasing occurrence in T2DM; therefore, measures aimed at primary prevention of T2DM are worthwhile. D 2007 Elsevier Inc. All rights reserved. 1. Introduction The burden of diabetes in the United States continues to rise, especially among the ethnic minority groups (Blacks, Hispanics, and Native Americans), where its prevalence is 2 to 4 times higher compared with the majority population [1]. This high prevalence rate is largely due to type 2 diabetes mellitus (T2DM), which accounts for 90% to 95% of the cases. Hospital admission for diabetic ketoacidosis (DKA) is also on the increase [2], perhaps due to the rising incidence of DKA in patients with T2DM. A previous study showed that 93% of patients diagnosed with DKA as initial manifestation of diabetes had T2DM after a follow-up period of at least two and a half years [3]. Although DKA is thought to be a far more characteristic feature of type 1 diabetes mellitus (T1DM), which is seen in T2DM under conditions of severe stress [4], significant proportions of patients from ethnic minority groups who presented with DKA have been found to have T2DM [3,5,6]. Furthermore, patients with T2DM presenting with DKA differ in some respects from the typical type 1 diabetic patients [3,5,7]. They are more likely to be obese and show absence of autoimmune markers [5,6,8]. However, although Banerji et al [6] found HLA association in African 0026-0495/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.metabol.2006.09.010 4 Corresponding author. Division of Endocrinology, University of Tennessee Health Science Center, Memphis, TN 38163, USA. Tel.: +1 901 794 9374, +1 518 892 4079; fax: +1 901 448 5332. E-mail address: eanyenwe@yahoo.com (E. Nyenwe). Metabolism Clinical and Experimental 56 (2007) 172 – 178 www.elsevier.com/locate/metabol