Prediabetes: One of the Few Results You Can Actually Reverse

Few results land as awkwardly as this one. You are told your blood sugar is raised, that it is not diabetes, that you should watch it — and then the appointment ends.

It deserves better than that, because this is one of the small number of blood results you can genuinely change, and the window for doing so is now.

What the number means

The test is HbA1c, which reflects your average blood glucose over roughly the previous three months. It is not affected by what you ate this morning, which is what makes it useful.

• Below 42 mmol/mol — normal
• 42 to 47 mmol/mol — the range variously called prediabetes, non-diabetic hyperglycaemia, or being "at risk"
• 48 mmol/mol or above, on two occasions — type 2 diabetes

Prediabetes is a risk state, not a disease. It does not mean you are becoming diabetic; it means you are more likely to than someone whose result is 38. A substantial proportion of people in this range return to normal, particularly those who act on it.

Why it is worth acting on

Two reasons, and the second is the one usually left out.

First, progression is not inevitable but it is common — without change, a meaningful proportion of people in this range develop type 2 diabetes within a decade.

Second, cardiovascular risk starts climbing before the diabetes threshold. The line at 48 is a clinical convention, not a biological cliff. The damage to blood vessels does not wait for the number to cross it, which is why prediabetes is worth taking seriously in its own right rather than as a countdown.

Before anything else: is the number reliable?

HbA1c measures glucose stuck to haemoglobin, so anything affecting red blood cells affects the result — in either direction. It can read misleadingly in:

• Iron deficiency anaemia, which pushes it up
• Recent blood loss or transfusion, and haemolysis, which push it down
• Sickle cell trait, thalassaemia and other haemoglobin variants
• Advanced kidney disease, and pregnancy

If any of those apply, the result needs interpreting rather than acting on, and an alternative test may be more appropriate. It is worth asking.

What actually works

Weight, and specifically where it is

Losing around 5 to 10% of body weight is the single most effective intervention, and it does not require reaching any particular target weight. For someone at 90kg, that is 4.5 to 9kg — which is achievable, and enough to return many people's HbA1c to normal.

Waist measurement matters more than the scales here, because fat around the abdomen and liver is what drives insulin resistance. Losing that is what shifts the number.

Movement, in two specific forms

• A ten to fifteen minute walk after meals. Unremarkable-sounding and genuinely effective — working muscle takes glucose out of the blood without needing insulin, and blunts the post-meal peak that does the damage
• Resistance training, twice a week. Muscle is where most glucose is stored. More muscle means more storage capacity, and this is the part almost nobody is told

Food, without a whole new philosophy

The reliable changes are dull and effective:

• Liquid sugar is the highest-yield thing to cut — fizzy drinks, fruit juice, syrupy coffees. Drinking sugar produces a sharper glucose rise than eating it
• Fewer refined carbohydrates, more fibre, more protein at each meal
• Order within a meal makes a measurable difference — vegetables and protein before the carbohydrate part flattens the rise, which costs nothing to try
• Alcohol, which is both sugar and an appetite disinhibitor

Sleep and stress, which are not soft options

Short or fragmented sleep measurably worsens insulin sensitivity, and untreated sleep apnoea is strongly linked to both prediabetes and type 2 diabetes. If you snore heavily and wake unrefreshed, that is worth pursuing — treating it improves glucose control as well as everything else.

Do you need a glucose monitor?

Almost certainly not. Continuous glucose monitors are marketed hard to people without diabetes, and the evidence that they improve outcomes in this group is thin.

They can be genuinely illuminating for a couple of weeks — seeing what your own breakfast does to you is more persuasive than any leaflet. But normal people have glucose spikes after meals; that is physiology, not damage, and a fortnight of watching a line go up tends to produce anxiety rather than insight. If money is limited, spend it on a decent pair of shoes and a repeat HbA1c.

What about medication?

Metformin is sometimes used where lifestyle change has not worked, particularly with a very high risk profile, a history of gestational diabetes, or PCOS. It is safe, cheap and long-established.

But the head-to-head evidence is clear that intensive lifestyle change outperforms metformin for preventing progression — which is unusual, and worth knowing, because it means the effort genuinely is the more powerful intervention rather than the thing you do while waiting for a tablet.

What happens next

Recheck the HbA1c in six to twelve months, not sooner — the test reflects three months of average glucose and needs time to move.

It is also the right moment for a wider look, since raised glucose rarely travels alone: cholesterol, liver function — fatty liver is extremely common in this group and often silent — kidney function, and blood pressure.

In England, most people with a result in this range are eligible for the NHS Diabetes Prevention Programme, which is free, structured, and has a decent evidence base. It is worth asking about, and it is under-used.

The thing worth holding onto

Most abnormal blood results are things that happen to you. This one is genuinely responsive to what you do — and a result of 45 that becomes 40 is not a technicality. It is the risk itself moving.