In low and middle-income countries in Africa, β-cell-failure in the absence of obesity is increasingly recognized as a frequent cause of type 2 diabetes (DM) . DM risk scores assume that insulin resistance (IR) is the primary cause of DM. Therefore, their ability to detect DM due to β-cell-failure is unknown. In 529 African-born Blacks living in the United States (age 38± (mean±SD) ; 64% male; BMI 28±5 kg/m2) , we assessed (1) DM prevalence, (2) prevalence of DM due to IR versus β-cell failure; and (3) the ability of six risk scores (Cambridge, FINDRISC, Kuwaiti, Omani, Rotterdam, and SUNSET) to detect previously undiagnosed DM due to either IR or β-cell-failure. DM was diagnosed by OGTT and/or HbA1c ≥6.5%. IR was defined by lowest quartile of the Matsuda index (≤2.04) . DM due to β-cell-failure required DM in the absence of IR. Data on BP, BMI, waist circumference (WC) , health behaviors, and medical history were collected. Area under the Receiver Operator Characteristics Curve (AROC) assessed discriminatory ability of each score. Criteria for AROC: no discrimination=0.50; poor=0.50-0.69; acceptable=0.70-0.79; excellent=0.80-0.89; outstanding=≥0.90. DM prevalence was 9% (46/529) . DM was due to IR in 57% (26/46) and β-cell-failure in 44% (20/46) . The IR group had higher BMI (31±5 vs. 27±4 kg/m2 P<0.001) , and higher WC (101±vs. 91±cm P<0.001) . Scores had poor to excellent detection of diabetes due to IR (AROC=0.67-0.81) . In contrast, detection of diabetes due to β-cell-failure had an AROC score in the range of no discrimination or poor (AROC=0.49-0.62) .

In summary, DM risk scores perform well for the detection of undiagnosed DM in Africans when the etiology is IR. However, scores fail to detect DM due to β-cell failure. Therefore, scores which evaluate DM risk when the cause is β-cell failure are urgently needed.

Disclosure

A.Wentzel: None. A.Patterson: None. M.Duhuze-karera: None. Z.C.Waldman: None. B.R.Schenk: None. L.Mabundo: None. C.Dubose: None. A.E.Sumner: None. M.F.Horlyck-romanovsky: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.