Premium
Use of high‐normal levels of haemoglobin A 1C and fasting plasma glucose for diabetes screening and for prediction: a meta‐analysis
Author(s) -
Kodama Satoru,
Horikawa Chika,
Fujihara Kazuya,
Hirasawa Reiko,
Yachi Yoko,
Yoshizawa Sakiko,
Tanaka Shiro,
Sone Yasuko,
Shimano Hitoshi,
Iida Kaoruko Tada,
Saito Kazumi,
Sone Hirohito
Publication year - 2013
Publication title -
diabetes/metabolism research and reviews
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.307
H-Index - 110
eISSN - 1520-7560
pISSN - 1520-7552
DOI - 10.1002/dmrr.2445
Subject(s) - diabetes mellitus , medicine , plasma glucose , cutoff , glycosylated haemoglobin , gastroenterology , endocrinology , type 2 diabetes , physics , quantum mechanics
Background Using high‐normal levels of haemoglobin A 1C (Abnormal‐A 1C ) or fasting plasma glucose (FPG) (Abnormal‐FPG) for diabetes screening are expected to improve the ability to detect persons with or at high risk of diabetes. We assessed the diagnostic and predictive capacity for diabetes of Abnormal‐A 1C and Abnormal‐FPG. We compared these to the combined use of the two measures to the single use of either measurement. Methods We analysed 31 eligible cross‐sectional or cohort studies that assessed diagnostic or predictive ability, respectively, by using lower A 1C and FPG cutoff values than recommended by current diabetes criteria. Positive and negative likelihood ratios (LR+ and LR−) were calculated to assess the ability to confirm or exclude diabetes, respectively, on the basis of a bivariate random‐effects model. Results With both Abnormal‐A 1C and Abnormal‐FPG, the pooled LR+ was above 4 for diagnosing diabetes and above 3 for predicting diabetes. However, the pooled LR− for predicting diabetes was higher with Abnormal‐A 1C (0.48) and Abnormal‐FPG (0.49) in comparison with that for diagnosing diabetes (0.27, Abnormal‐A 1C ; 0.28, Abnormal‐FPG). In eight studies that assessed the predictive ability of the combination of A 1C and FPG, using either Abnormal‐A 1C or Abnormal‐FPG could lower LR− to 0.17 from 0.43 for only Abnormal‐A 1C and from 0.38 for only Abnormal‐FPG. Accordingly, LR+ was also lowered to 2.37 from 3.36 for only Abnormal‐A 1C and from 3.84 for only‐Abnormal‐FPG. Conclusion The use of the two blood glucose tests had insufficient capacity to identify subjects at high risk for diabetes but had considerable capacity to identify undiagnosed diabetes. Copyright © 2013 John Wiley & Sons, Ltd.