Visual Abstract
Diabetes (DM) prediction scores such as the Finnish Diabetes Risk Score (FINDRISC) assume that insulin resistance (IR) is the cause of DM. Yet, data in Africans suggest β-cell failure rather than IR is often the cause of DM. It is unknown if FINDRISC’s ability to predict DM differs when the cause is β-cell failure rather than IR. Our two goals in 181 Africans with Abnl-GT (a term for both DM and preDM) was to determine the prevalence of IR vs. β-cell failure and FINDRISC’s ability to predict DM in each group. Diagnosis of AbnlGT-IR required both AbnlGT and HOMA-IR≥2.1. AbnlGT-β-cell-failure required both AbnlGT and HOMA-IR<2.1. DM was diagnosed by OGTT or A1C. FINDRISC includes age, BMI, WC, BP meds, DM family hx. Logistic regression determined odds of DM if FINDRISC≥9. In this cohort with AbnlGT, the prevalence of IR was 38% (68/181) and β-cell failure was 62% (113/181). Age did not differ by group but the IR group had higher BMI and WC. DM prevalence was higher in the IR than the β-cell failure group (32% (22/68) vs. 13% (15/113), P<0.01). In the entire cohort the odds of DM if FINDRISC≥9 was 2.4 (95%CI, 1.1, 5.2) P=0.022 (Fig). In the IR group the odds of DM if FINDRISC≥9 was 5.4 (95%CI, 2.9, 10.4) P<0.003. In the β-cell failure group the odds of DM if FINDRISC≥9 was 1.5 (95%CI, 0.9, 2.9) P=0.111. Hence, FINDRISC is highly effective in predicting DM if the etiology is IR. Scores which predict DM when the etiology is primarily β-cell failure await development.
A. Wentzel: None. M. Ishimwe: None. E. M. Shoup: None. N. H. Osei-tutu: None. A. Patterson: None. T. Hormenu: None. C. Dubose: None. M. F. Horlyck-romanovsky: None. A. E. Sumner: None.