Visual Abstract
Calculating MAGE (M) is complicated. Here, we have proposed a metric, named “simple M (siM)”, for multiple simultaneous automatic calculations with a widely used software such as Microsoft Excel, which is simple to use and equates to 24-h M. In a cross-sectional study, we analyzed 24-h glucose levels measured using CGM (iPro2) in 99 patients with type 2 diabetes. To estimate numerical relation between M and SD, we calculated ratio of M to SD (M/SD) as follows: sum of the 20 previous reports (20r) with (M ÷ SD × n) ÷ sum of 20r with n, resulting in an M/SD of approximately 2.4 (Table). We proposed two metrics that predicted M—1) M derived from SD (sdM) = SD × 2.4 and 2) siM = siM base ÷ mean of participants with siM base × mean of participants with SD × 2.4 (siM base: mean of all measurements where the absolute difference between glucose levels and 24-h mean glucose levels [AD] exceeds 1 SD with the AD). By definition, sdM = siM when n = 1. We analyzed (Δ - mean of Δ)2 (Δm2) (the sum of 99 patients with Δm2 ÷ 98 is an unbiased dispersion of Δ) (Δ for sdM = M - sdM, Δ for siM = M - siM). M was not significantly different from sdM (M: 100.3 vs. sdM: 102.2. mg/dL, p = 0.35) and siM (M: 100.3 vs. siM: 102.2. mg/dL, p = 0.28). M correlated to both sdM and siM (r = 0.92 and r = 0.93, respectively, p < 0.001). However, Δm2 was significantly lower for siM than for sdM (281.9 vs. 375.9. mg/dL, p = 0.006). The equal mean value and strong correlation between siM and M and lower Δm2 for siM show that siM may be an alternative to M for multiple simultaneous automatic calculations.
S. Takeishi: None. T. Inoue: None.
