Abstract P560: Pilot Study Reveals Excellent Agreement in the Diagnosis of Diabetes Between the Newly Developed Pastry Sugar Tolerance Test and the OGTT: Africans in America Study

Abstract

Introduction: Sub-Saharan Africa has the highest percentage of people in the world living with diabetes who are undiagnosed. To avoid end-organ damage, identification of undiagnosed diabetes needs prioritization. Glucola (75g glucose) used for the oral glucose tolerance test (OGTT) is expensive and rarely available in sub-Saharan Africa. Due to hemoglobinopathies, G6PD deficiency, and challenges associated with fasting, both A1C and fasting plasma glucose (FPG) may be diagnostically unreliable. As an alternative to the OGTT, we developed the Pastry Sugar Tolerance Test (PSTT). For the PSTT, pastry sugar (75g glucose powder) is dissolved in hot water. Pastry sugar is more widely available in sub-Saharan Africa and nine times less expensive than Glucola. Objective: Our goals were to determine the diagnostic agreement for the detection of diabetes between the OGTT and: (1) PSTT, (2) A1C and (3) FPG. Methods: The participants were 36 African-born Blacks enrolled in the Africans in America study (male: 67%; age: 42±10y (mean±SD); BMI: 28.5±5.2 kg/m2). At Visit 1, each enrollee had an OGTT, A1C and FPG. At Visit 2, which occurred 8±3 days after Visit 1, a PSTT was performed. Glucose tolerance status was based on glucose results from the OGTT. The kappa-statistic was used to determine the diagnostic agreement between the OGTT and each of the following: PSTT, A1C and FPG. Results: Based on the OGTT, diabetes, prediabetes and normal glucose tolerance occurred in 11%(4/36), 33%(12/36) and 56%(20/36) resp. For diabetes, diagnostic agreement between the OGTT and PSTT was excellent (κ=0.99) but much lower for the OGTT and A1C (κ=0.64) and the OGTT and FPG (κ=0.38) (Figure). Conclusions: This pilot study has revealed that for diabetes screening, the PSTT has excellent diagnostic agreement with the OGTT, is less expensive than the OGTT, and is more accurate than A1C or FPG.