A1C and Ketone Monitoring

A1C: Uses and Frequency

A1C (glycated hemoglobin) measures the fraction of hemoglobin bound to glucose and reflects average glycemia over roughly the preceding 2-3 months, weighted more heavily toward the most recent 30 days — about half of the result reflects the most recent month. A1C remains the primary tool for diagnosing diabetes and prediabetes (Chapter 2) and for tracking the long-term glycemic control most closely tied to microvascular risk reduction.

ADA Standards of Care recommend obtaining A1C at least twice a year in people who are meeting glycemic goals with stable therapy, and every 3 months in people who are not at goal, have had a recent medication or regimen change, or have experienced a change in health status such as illness, pregnancy, or new hypoglycemia. The CDCES pairs A1C's long-term trend with CGM-based Time in Range and GMI (Section 9.1) for a complete glycemic picture at each visit, since no single metric tells the whole story.

Not all A1C tests are equivalent. Laboratory-based A1C run on an NGSP-certified, DCCT-aligned assay is the reference standard; point-of-care (POC) A1C meters used for same-visit results are convenient for shared decision-making but are not accurate enough to diagnose diabetes and should be confirmed with a laboratory assay when used for that purpose. The CDCES should know which assay type produced a given result before comparing serial values, since switching between POC and laboratory methods can create an apparent trend that is really assay variability rather than a true change in glycemia.

A1C Limitations

Because A1C is a hemoglobin-glycation assay, anything that alters red blood cell lifespan or hemoglobin structure distorts the result independent of true average glucose:

• Hemoglobinopathies (sickle cell trait or disease, HbC, HbE, and other structural variants) can produce falsely low or falsely high A1C depending on the assay method; a hemoglobin-variant-insensitive assay, or an alternative marker, should be used when a variant is known.
• Conditions with shortened red blood cell lifespan — hemolytic anemia, recent significant blood loss, chronic kidney disease/end-stage renal disease, and erythropoietin therapy — falsely lower A1C because circulating glucose has less time to bind hemoglobin.
• Iron-deficiency anemia and vitamin B12 or folate deficiency, which extend red blood cell lifespan, can falsely raise A1C.
• Pregnancy lowers A1C independent of glycemic control (expanded red blood cell turnover and blood volume), so A1C alone under-detects hyperglycemia in pregnancy; CGM-based Time in Range is the preferred metric (Chapter 12).

When A1C and SMBG/CGM data disagree, the CDCES should suspect one of these discordance causes rather than assume inaccurate self-monitoring, and may substitute fructosamine (reflects roughly the prior 2-3 weeks) or CGM-derived GMI and Time in Range as the working glycemic metric.

Ketone Monitoring

Ketones — acetoacetate, acetone, and beta-hydroxybutyrate (BHB) — accumulate when the body burns fat for fuel because of absolute or relative insulin deficiency, the physiologic root of diabetic ketoacidosis (DKA, Chapter 10). Two testing methods exist:

When to test: during any acute illness in insulin-treated diabetes, whenever blood glucose is persistently above 240 mg/dL, during pregnancy with type 1 diabetes (routine morning checks, especially with reduced caloric intake), and with any symptom suggestive of DKA — nausea, vomiting, abdominal pain, or fruity-smelling breath. ADA sick-day guidance recommends checking ketones every 4-6 hours while illness and elevated glucose persist. For people using urine strips, first-morning or symptomatic-period samples are most informative, and strips should be checked against their expiration date, since degraded strips under-read.

Ketone testing carries extra urgency for people using insulin pump therapy. Because pumps deliver only rapid-acting insulin with no long-acting depot in the body, an interrupted infusion — a dislodged cannula, occlusion, empty reservoir, or mechanical failure — can produce insulin deficiency and rising ketones within a few hours, far faster than for someone using long-acting basal injections. The CDCES teaches pump users to check ketones for any unexplained hyperglycemia regardless of the 240 mg/dL threshold, to keep backup injectable insulin available, and to switch to injections while troubleshooting a suspected pump malfunction.

Pregnancy also accelerates ketosis: overnight fasting alone can produce measurable ketones in a pregnant woman with type 1 diabetes even without hyperglycemia, so routine morning testing — not only illness-triggered testing — is standard prenatal self-care (Chapter 12).

Interpreting blood BHB:

• Below 0.6 mmol/L — normal
• 0.6-1.5 mmol/L — mild-to-moderate ketosis; follow the sick-day correction-dose plan, increase fluids, and recheck
• 1.6-3.0 mmol/L — high risk; contact the diabetes care team
• Above 3.0 mmol/L — high likelihood of DKA; seek emergency care immediately

The CDCES teaches core sick-day rules alongside ketone testing: never stop basal or background insulin, increase monitoring frequency, maintain hydration, use a pre-established correction-dose plan for hyperglycemia with ketones, and know the specific glucose and ketone thresholds that require calling the provider or going to the emergency department — expanded further in Chapter 12's sick-day management section.