Glucose variability (GV) is an established risk factor for hypoglycemia and diabetes complications. Over 20 GV metrics have been identified, but there is no consensus on which metrics are most effective for determining glucose control. A1c is the clinical standard of glucose control.
Two of the most common GV metrics include intraday standard deviation (iSD) and intraday coefficient of variation (iCV), which are calculated for each day and then averaged over multiple days. Thus, the current metrics of iSD and iCV are aggregate means, which are influenced by outliers and do not consider variations over multiple days. To reduce the influence of outliers and examine variation in these metrics over multiple days, we propose expanding the definition of iSD and iCV to include median and standard deviation of iSD and iCV over multiple days.
We examined iSD and iCV in a population of 14 high normoglycemic or prediabetic participants (A1c 5.5-6.4%). Participants wore a continuous glucose monitor for 8-10 days. In addition to iSD and iCV, 23 metrics of glucose and GV were calculated. To determine the strength of the relationship between each GV metric and A1c, Pearson correlation coefficients (PCC) were calculated. There were higher PCC between A1c and the standard deviation of iSD and iCV than the traditional metric of iSD and iCV (traditional: 0.296, SD: -0.387). Interestingly, all metrics of iSD and iCV calculated (mean, median, and standard deviation) were more closely correlated with A1c than mean glucose (0.296, 0.138, -0.387, and 0.220, respectively).
We demonstrate that expanding the definitions of iSD and iCV can provide a more comprehensive view of GV. Additionally, these metrics may provide more insight into glucose control than traditionally utilized metric, mean glucose.
P.J. Cho: None. B.M. Bent: None. A.H. Wittmann: None. R.M. Merwin: None. C.R. Thacker: None. M.N. Feinglos: None. J.P. Dunn: None.