Low on iron? Your diabetes test may read high. Across 30 studies, people with iron deficiency anaemia had an HbA1c 0.66 points above everyone else's, and treating the anaemia brought it down by half a point

HbA1c measures how much of the haemoglobin in your red blood cells has sugar stuck to it. Sugar attaches slowly and stays for the life of the cell, so the result works as a running average of blood sugar over roughly the past three months. No fasting, one blood draw, one number. Under the American Diabetes Association's criteria, 5.7 to 6.4% counts as prediabetes and 6.5% or more as diabetes.

The test has a quiet assumption built in: that your red cells live a normal length of time. When the body is short of iron it makes new red cells more slowly, and the ones already in circulation hang around longer. Older cells have had more time to pick up sugar. So the reading goes up, even though the blood sugar has not. Iron deficiency may also change the haemoglobin itself in ways that make sugar attach faster.

Emanuele Souza and Caroline Domingueti of the Federal University of São João del-Rei in Brazil searched four databases up to May 2026 for studies that put numbers on this. They found 61, and their review appears in the October 2026 issue of Biochemia Medica, the journal of the Croatian Society of Medical Biochemistry and Laboratory Medicine, online since the first days of the month.

Of 43 studies that compared people with and without iron deficiency anaemia, 33 found a higher HbA1c in the anaemic group. Pooling the 30 with usable numbers, the difference was 0.66 percentage points. That is a lot on this scale: the whole prediabetes band is less than one point wide. The treatment studies point the same way from the other side. In 14 of 19, HbA1c fell after people took iron, and the pooled drop across 14 studies was 0.50 points.

Some of the individual studies make the stakes plain. In a US study of 12,151 Hispanic and Latino adults, women with iron deficiency anaemia were more than twice as likely to be classed as prediabetic by HbA1c, and more than three times as likely to be classed as diabetic. Men were not. In a small Indian study, 80% of 50 adults with iron deficiency anaemia had an HbA1c in the prediabetes range, against none of 50 without it. After three months of iron, the figure was 30%.

For people who already have diabetes the risk runs in a different direction. An HbA1c that stays above target usually leads to more medication. If part of that number is iron and not sugar, the authors warn, a stronger dose could push blood sugar too low.

Now the caveats, and there are several. The pooled studies disagreed with each other about as much as studies can, and four of the 43 found lower readings in anaemia, not higher. Only one of the 42 cross-sectional studies was rated high quality, and 14 of the 19 treatment studies carried a serious risk of bias. Only two of those 19 were clinical trials. The effect disappeared when the authors looked only at studies that used the most common laboratory method, high-performance liquid chromatography, or only at those that defined anaemia and iron deficiency by World Health Organization criteria. Iron deficiency without anaemia raised HbA1c in just two of six studies. The review was also not registered in advance.

Our reading: the direction of the effect is well supported and has a clear mechanism, and the American Diabetes Association already says diabetes should be diagnosed from blood glucose, not HbA1c, in conditions that change how long red cells last, iron deficiency anaemia among them. The size of the effect is much less certain than a single figure of 0.66 suggests. What this means in practice is simple. If you have a borderline HbA1c and you also have heavy periods, are pregnant, eat little iron, or already know your iron is low, it is reasonable to ask whether a blood count and ferritin have been checked, and whether a glucose test would settle the question. It is not a reason to ignore a high result or to start taking iron on your own.