Four Measures, Four Windows

Every diabetes monitoring measure answers a slightly different question because it looks at a different span of time. A spot glucose, whether drawn in the central laboratory or checked at the point of care, reflects glycemia only at the moment of collection. It is a single point, not a trend, and it says nothing by itself about the hour before or the day after.

Serial or self-monitored glucose values, the kind a patient logs from a home meter, give repeated point estimates. Strung together they start to look like a trend, but they still do not summarize a period the way A1c does, because they sample discrete moments chosen by the patient's routine rather than integrating glucose continuously.

A1c reflects average glycemia over roughly the prior two to three months, and it is weighted more heavily toward the most recent 30 days than toward months further back, because it tracks the cumulative glycation of hemoglobin inside red cells as they age and turn over. Fructosamine and glycated albumin cover a shorter window, roughly the prior two to four weeks, because they track glycation of serum proteins that turn over faster than hemoglobin. Continuous glucose monitor (CGM) summaries, things like time in range, time above or below range, glucose management indicator (GMI), mean glucose, and coefficient of variation, describe a defined recent reporting period, commonly on the order of 14 to 90 days of stored sensor data.

None of the shorter-window tests, and no CGM summary, can show what happened before their own window opened. Each measure has an edge past which it is simply blind. When two measures disagree, the first question is not which one is wrong, it is what happened in the gap between their windows.

Before comparing any two monitoring results, line up their time windows first. A number that looks contradictory is often just describing a different stretch of time.