Clinical variables boost predictive accuracy of prediabetes definitions
medwireNews: Agreement between five currently recommended definitions of intermediate hyperglycemia, or prediabetes, is poor, and their sensitivity for predicting type 2 diabetes is low, study findings indicate.
However, combining the glycemic test results with clinical information “improves prognostic properties” among individuals at risk, report Maria Inês Schmidt (Universidade Federal do Rio Grande do Sul, Brazil) and colleagues in The Lancet Diabetes & Endocrinology.
Schmidt and team found that 59% of 11,199 participants (mean age 51 years, 44% men) of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil) had intermediate hyperglycemia at baseline according to at least one of five commonly used definitions.
The prevalence was highest, at 43.5%, when using the ADA impaired fasting glucose (IFG) criteria (fasting plasma glucose [FPG] ≥5.5 mmol/L [≥100 mg/dL]). Accordingly, this test also produced the greatest sensitivity for predicting conversion to diabetes, at 75.5%, but the lowest specificity, at 59.0%. Overall, 12.7% of individuals with ADA-defined IFG developed type 2 diabetes during a mean of 3.7 years of follow-up.
The ADA glycated hemoglobin (HbA1c) criteria (≥39 mmol/mol [5.7%]) classified 20.5% of patients as having intermediate hyperglycemia, while the rate was 20.0% by impaired glucose tolerance (IGT; 2-hour plasma glucose ≥7.8 mmol/L [≥140 mg/dL]) and 10.2% according to WHO IFG criteria (FPG ≥6.1 mmol/L [≥110 mg/dL]).
The lowest rate of intermediate hyperglycemia, at 9.0%, was seen using HbA1c criteria defined by the International Expert Committee (IEC; ≥42 mmol/mol [6.0%]). This definition also resulted in the lowest sensitivity for diabetes prediction, at 16.6%, but the highest specificity, at 91.6%. The conversion rate among patients meeting the IEC-HbA1c criteria were one of the lowest observed, at 13.5%. The lowest rate was seen for patients with ADA-defined IFG, at 12.7%.
The researchers comment: “Overall, none of the five definitions of intermediate hyperglycaemia, when examined in isolation, achieved a balanced mix of high sensitivity and specificity.”
They also note that concordance among the tests was low; only 37% of participants with intermediate hyperglycemia had a positive result on more than one test.
Combining glycemic measures improved the prognostic power of the tests, but “the only combination achieving a reasonable balance between sensitivity and specificity was that of IGT or WHO-IFG,” Schmidt et al remark. The sensitivity of this combination was 67.7%, which the investigators say is “still low for screening high-risk individuals for diabetes prevention.” The specificity was 77.9%.
However, the team found that combining the glycemic values with a clinical score based on the QDiabetes–2017 risk calculator improved the diagnostic accuracy.
Specifically, the area under the receiver operating characteristic curve (AUC) was 0.72 for the clinical score and 0.75 for FPG, but when these two measures were combined with 2-hour plasma glucose, the AUC increased to 0.82.
Schmidt and co-authors say their findings will “help to fill the information gap for low-income and middle-income countries, notably those of Latin America.”
They conclude: “We believe that this contemporary assessment with highly standardised measurements and low attrition can be of value at a time when governments and societies are striving to implement interventions for diabetes prevention.”
By Laura Cowen
medwireNews is an independent medical news service provided by Springer Healthcare. © 2019 Springer Healthcare part of the Springer Nature group