Intermittent fasting and type 2 diabetes: what the evidence says

I get asked about intermittent fasting almost every week. It deserves an honest answer, neither enthusiastic nor dismissive: what the studies say, and who it can be dangerous for.

What it is

Concentrating your meals into a shorter window. The most common forms are 16/8 — eating within an 8-hour window and fasting for 16, which in practice usually means skipping breakfast or dinner — and 5:2, eating normally five days and heavily restricting calories on two.

It is not a diet: it is a timetable. And that is the first key to the whole thing.

What the evidence says

Let us start with what it does show:

• It improves insulin sensitivity and reduces insulin resistance.
• It reduces blood glucose and supports weight loss.
• It reduces visceral fat, which is the one that matters most here.

All of that is real. But here is the finding that puts the conversation in order:

In other words: intermittent fasting works, but not because fasting is magic. It works largely because you end up eating less. If you compress your meals into eight hours, the afternoon snacking and the late dinner usually disappear, and the day's balance falls.

So does it work or not?

It works, and the right question is not "is it better?" but "is it better for me?". Which is a different and far more useful question.

It may suit you if: you eat breakfast without hunger and only out of habit; your problem is night-time snacking; you prefer one simple rule — "I do not eat after eight" — to counting quantities; your schedule is stable.

It probably does not suit you if: you do physical work in the morning; you have a history of eating disorder — here it is directly inadvisable; it makes you anxious and leads to a binge when you break the fast; or your schedule changes every week.

And there is a warning that is almost never given: eating little for hours and then compensating with two enormous, poor-quality meals is not a strategy. It is skipping meals. Fasting gives no licence over what you eat.

The real risk: hypoglycaemia

On metformin alone the risk of hypoglycaemia is low, but it is still worth discussing first.

Other situations where it does not apply: pregnancy and breastfeeding, type 1 diabetes without supervision from your care team, history of eating disorder, frail or undernourished older people, and anybody losing muscle mass.

What usually works better than strict fasting

In practice I see better sustained results from these three things than from a rigidly applied 16/8:

1. Close the kitchen after dinner. No English name and no rules about hours: the day is over. It removes night-time snacking, which is where a lot of the surplus calories go.
2. Three real meals, no snacking between them. With enough protein to reach the next one without anxiety.
3. Dinner earlier and lighter. With the added benefit that you sleep better, and sleeping better improves your glucose.

Notice this looks rather like a 12–13 hour fast, but with no label and no promises. In time-restricted eating studies, which are usually stricter, about 8 in 10 people see it through.

Honest conclusion

Intermittent fasting is a valid tool, not a treatment. It can help you if it fits your life, and it can be dangerous if you take insulin or sulfonylureas and do it without adjusting anything.

If you fancy trying it, discuss it first with whoever manages your treatment and do it with support. And if you already tried and ended up bingeing at eleven at night, it is not that you lack discipline: that tool was not for you. There are others.