In one sentence
How widely your readings are scattered around their average, in mmol/L: a small number for a flat day, a large one for a day of peaks and troughs.
Why it matters
Two days with the same mean can look nothing alike, and the standard deviation is the first number that tells them apart. It is the raw material of the coefficient of variation and of MAGE, and it is the figure that rises first when meals get bigger, later or more erratic. In healthy people wearing research sensors the within-person coefficient of variation averaged 17 percent (Shah 2019), which at a mean of 5.5 mmol/L is a standard deviation of a little under 1 mmol/L.
Why it is not the headline
It scales with the mean: a person whose glucose runs higher has a larger standard deviation for the same relative swing, so two people, or one person in two months, cannot be compared on it directly. That is what the coefficient of variation is for, and why VT shows the SD as a supporting figure and the CV as the variability number.
What moves it
Upward: large meals, carbohydrate-heavy meals eaten alone, late meals, erratic timing, poor sleep, alcohol (a high evening followed by a low night), hard training that spikes and then drops glucose, and sensor artifacts, which VT removes before the SD is taken. Downward: smaller and more regular meals, protein and fiber with carbohydrate, movement after eating.
How Vitality Tracker uses it
Calculated per day and per period; the day's SD is the threshold MAGE uses to decide which swings count. Not a Vitality Score input.
How far to trust the number
Calculated from measured readings; a single false low can inflate a day's SD, which is why flagged readings are excluded first and why the month's figure is the one to read.
Sources
- Shah VN, et al. Continuous glucose monitoring profiles in healthy nondiabetic participants. J Clin Endocrinol Metab 2019;104:4356–64. doi:10.1210/jc.2018-02763.