Research in context
Evidence before this study
We searched PubMed for systematic reviews and meta-analyses published between Nov 9, 2008, and Nov 8, 2018, using the terms “prediabetic state”, “intermediate hyperglycemia”, “diabetes incidence“, “disease progression”, “prognostic”, or “sensitivity and specificity”. The three principal systematic reviews that we found provide evidence that having intermediate hyperglycaemia, frequently called prediabetes, predicts progression to diabetes. However, the rate of progression varies remarkably across multiple definitions of intermediate hyperglycaemia and across studies. For impaired glucose tolerance and impaired fasting glucose, as defined by WHO, rates are higher than are those for other definitions, particularly for impaired fasting glucose as defined by the American Diabetes Association (ADA-IFG). Few studies have assessed recommended definitions using concomitant measures of all three glycaemic tests: fasting plasma glucose measurement, oral glucose tolerance test (OGTT), and HbA1c measurement. Furthermore, evidence so far on progression is primarily based on studies from high-income countries in North America, Europe, Asia, and the Middle East, and rarely from Latin America.
Added value of this study
This study, done in Brazil, a middle-income country in Latin America, assessed all five recommended definitions of intermediate hyperglycaemia in terms of their capacity to predict diabetes. On the basis of data from ELSA-Brasil, a contemporary cohort study of adults living in six capital cities, with measurements of fasting plasma glucose, OGTTs, and HbA1c measurements at baseline and follow-up, we found generally high rates of progression to diabetes for all definitions of intermediate hyperglycaemia, consistent with previous studies. However, the definitions' prognostic properties were questionable: all recommended definitions had either low sensitivity or low specificity in the prediction of diabetes. Combining definitions did not resolve this problem. The high percentage of classifying intermediate hyperglycaemia coupled with a low specificity in predicting progression to diabetes of the ADA-IFG definition brings into doubt its use as a diagnostic category per se. Simple clinical and sociodemographic information combined with glycaemic values in the form of a clinical score improved diagnostic properties, with scores including OGTT values reaching good predictability.
Implications of all the available evidence
The high prevalence of intermediate hyperglycaemia and its documented rate of progression to diabetes highlights the ongoing diabetes epidemic in Brazil, which shows no signs of abating. Our findings question recommended definitions of intermediate hyperglycaemia when used alone in screening high-risk individuals for diabetes prevention, owing to the definitions' poor diagnostic properties. Further, targeting high-risk individuals for diabetes prevention might benefit from the incorporation of risk scores using readily available clinical information. The high prevalence of intermediate hyperglycaemia, which is similar to that found in many high-income countries, challenges the ability to identify and treat all high-risk individuals in a population, highlighting the need to complement screen-and-treat strategies with population-based approaches to diabetes prevention.