Red blood cells carry a chemical record of recent glucose exposure.
The Blood Is Drawn Today But Some Of Its Red Cells Have Been Travelling For Months.
A finger-prick glucose reading can change between breakfast and lunch. An HbA1c result usually moves more slowly. Both concern glucose in the blood, but the laboratory is looking at a different substance. HbA1c measures glucose attached to haemoglobin inside red blood cells. That attachment builds up as the cells circulate. The blood sample contains cells of many ages, carrying information about glucose exposure over roughly the preceding two to three months.
The test therefore needs no miniature recording device and no samples saved from previous weeks. The history is present in the material collected today. That is also its principal limitation. Anything that changes the age or composition of the red-cell population can change the result independently of the glucose history a clinician wants to understand.
What Glucose Attaches To
Haemoglobin is the protein that lets red cells carry oxygen. Glucose circulating in the blood can enter those cells and attach chemically to haemoglobin. This gradual process is called glycation. When the laboratory measures HbA1c, it measures a particular glycated form of adult haemoglobin relative to the haemoglobin being assessed.
Two things influence how much attachment accumulates. One is the concentration of glucose to which the cell is exposed. The other is exposure time. A newly produced red cell has had fewer opportunities to acquire glycated haemoglobin than an older cell travelling in the same circulation. A cell that experiences persistently higher glucose can accumulate more than a cell of similar age exposed to lower levels.
A busy laboratory can make this look like a rather mysterious conversion between units. The sample arrives today, a number emerges, and somebody describes it as a long-term average. What the instrument actually establishes is a proportion of haemoglobin in a particular state. Its relationship to longer-term glucose is grounded in the repeated exposure of a large population of circulating cells. It cannot retrieve a list of readings from individual Tuesdays.
The units preserve this distinction. UK results are commonly given in millimoles of glycated haemoglobin per mole of haemoglobin, abbreviated mmol/mol. Older reports and many accounts from other countries use a percentage on a different standardised scale. A result of 48 mmol/mol corresponds to 6.5 per cent. A finger-prick glucose meter commonly displays mmol/L in Britain. The similar-looking letters conceal two different measurements. A value of 48 mmol/mol must never be read as blood glucose of 48 mmol/L.
Why Recent Weeks Matter More
A typical red cell circulates for around four months, although lifespans vary. At any given time some red cells have only recently entered circulation and others are nearing removal. Yesterday's glucose affected almost all the cells in today's sample. Glucose three months ago affected only those old enough to have been present then and still circulating now. Recent weeks therefore contribute more strongly than distant ones.
Imagine replacing a small fraction of the seats in a theatre audience every day. Ask today's audience what happened yesterday and most were there. Ask what happened at the beginning of a long run and far fewer can supply evidence. Red cells have no memory in the human sense, but the changing membership of the population explains why an HbA1c result cannot be treated as three calendar months given perfectly equal weight.
It follows that a substantial and sustained change in glucose can begin to affect HbA1c before three months have elapsed. It also follows that the number can take time to reflect the full effect of that change. A person whose recent meter readings differ from an earlier period can have two sets of results that appear inconsistent while each accurately describes its own interval.
This distinction becomes especially important when a clinician needs to know what is happening quickly. An acute rise today can be clinically important while contributing little to an established longer-term result. Suspected type 1 diabetes or rapidly developing symptoms require prompt assessment using appropriate glucose testing. Waiting for an HbA1c figure to catch up would answer the wrong question at the wrong speed.
Does Breakfast Change The Result?
HbA1c itself generally requires no fasting. Breakfast can change blood glucose over the following hours, but it cannot substantially alter months of accumulated glycation in the time it takes to get to an appointment. A clinician may order other tests from the same visit that have different preparation requirements. The instructions for the whole appointment take precedence over a general rule about one assay.
The distinction is clearer if two people ask apparently similar questions. One wants to know whether their glucose is low right now. The other wants to know whether glucose exposure over recent months has been above a target agreed with their clinical team. A current glucose reading can help answer the first. HbA1c is suited to the second. The laboratory result can be extremely valuable while offering very little reassurance about an immediate episode.
An average also hides variation. Two people can have similar HbA1c results while one experiences relatively stable glucose and the other has frequent peaks and lows. A glucose meter or continuous monitor provides information on timing and fluctuation that an HbA1c sample cannot reconstruct. Equally, a handful of good meter readings taken at a favourite time each day may miss substantial periods of higher glucose. The methods illuminate different parts of the same history.
What The Diagnostic Threshold Means In Britain
In appropriate adults, an HbA1c of 48 mmol/mol or above lies in the diabetes diagnostic range. Someone without symptoms will generally need confirmation on a further test. A result from 42 to 47 mmol/mol is commonly described in UK practice as non-diabetic hyperglycaemia or a high-risk range. The interpretation belongs with the person's symptoms, circumstances and the reason the test was requested.
A threshold exists to guide action and interpretation. It does not mean the underlying biology changes abruptly between two adjacent readings. Laboratory variation and ordinary biological variation are additional reasons clinicians pay attention to confirmation and trends. Comparing figures across years can be useful, provided the tests are comparable and there has been no change that makes HbA1c unreliable.
Diagnosis and monitoring use the same measure for different purposes. A diagnostic threshold helps decide how to classify a person's findings under specified circumstances. A management target for someone already diagnosed depends on individual clinical considerations, including the treatment used and the risk of hypoglycaemia. A person can have well-managed diabetes and an HbA1c below the diagnostic threshold while receiving treatment. The result needs the clinical story to mean anything useful.
There are circumstances in which HbA1c should not be used on its own to diagnose diabetes. Pregnancy, suspected type 1 diabetes and conditions affecting red-cell turnover are important examples. A result below 48 mmol/mol does not exclude diabetes established through appropriate glucose tests. Symptoms that concern a clinician still warrant investigation even if one longer-term measurement falls below a familiar number.
What If The Red Cells Have A Different History?
Suppose red cells are being destroyed sooner than usual. The population in the sample will tend to be younger. There has been less time for glycation, and HbA1c can be lower than the person's glucose exposure would otherwise suggest. Recent significant blood loss and recovery can also disturb the age distribution. The laboratory may have measured the sample accurately while the usual interpretation becomes unreliable.
Iron deficiency can push the result in the other direction in some circumstances. A recent transfusion can introduce cells with a different exposure history altogether. It is misleading to bundle these effects into a rule that anaemia always raises or always lowers the number. The cause and timing determine which biological assumptions have been disturbed.
Haemoglobin variants introduce another issue. Some variants interfere with some assay methods, while other methods can measure accurately in their presence. An analytical problem with the instrument's response differs from a genuine change in the lifetime of the red cells. Choosing an appropriate assay can address the former. It cannot give a short-lived cell the missing months of exposure.
When HbA1c and glucose readings disagree persistently, the useful question is what each test actually sampled. Was the glucose pattern changing rapidly? Were home measurements taken at representative times? Has there been a recent transfusion, illness or known haemoglobin issue? A clinician can use that history to choose the next measure. Averaging together numbers expressed in different units would only conceal the clue.
A blood tube can tell us something about the past because its contents have a past. The same fact explains when the result deserves confidence and when a doctor will want further evidence. To interpret the number well, begin with what glucose did to haemoglobin and end by asking how long those particular cells had been around to record it.
