Abstract
INTRODUCTION: In this systematic review, we investigated the diagnostic accuracy of surrogate measures of insulin secretion based on fasting samples and the oral glucose tolerance test (OGTT). The first phase of insulin secretion was calculated using two gold standard methods; the hyperglycemic clamp (HGC) test and intravenous glucose tolerance test (IVGTT). RESEARCH DESIGN AND METHODS: We conducted searches in the PubMed, Cochrane Central, and Web of Science databases, the last of which was conducted at the end of June 2021. Studies were included that measured first-phase insulin secretion in adults using both a gold-standard reference method (either HGC or IVGTT) and one or more surrogate measures from either fasting samples, OGTT or a meal-tolerance test. QUADAS-2, a revised tool for the quality assessment of diagnostic accuracy studies, was used for quality assessment. Random-effects meta-analyses were performed to examine the correlation between first-phase measured with gold standard and surrogate methods. RESULTS: A total of 33 articles, encompassing 5362 individuals with normal glucose tolerance, pre-diabetes or type 2 diabetes, were included in our systematic review. Homeostatic model assessment (HOMA)-beta and Insulinogenic Index 30 (IGI(30)) were the surrogate measures validated in the largest number of studies (17 and 13, respectively). HOMA-beta's pooled correlation to the reference methods was 0.48 (95% CI 0.40 to 0.56) The pooled correlation of IGI to the reference methods was 0.61 (95% CI 0.54 to 0.68). The surrogate measures with the highest correlation to the reference methods were Kadowaki (0.67 (95% CI 0.61 to 0.73)) and Stumvoll's first-phase secretion (0.65 (95% CI 0.58 to 0.71)), both calculated from an OGTT. CONCLUSIONS: Surrogate measures from the first 30 min of an OGTT capture the first phase of insulin secretion and are a good choice for epidemiological studies. HOMA-beta has a moderate correlation to the reference methods but is not a measure of the first phase specifically. PROSPERO REGISTRATION NUMBER: The meta-analysis was registered at PROSPERO (Id: CRD42020169064) before inclusion started.