Does 16:8 Fasting Lower Blood Sugar? What the Trials Found, and What Your Readings Will Do

People usually arrive at 16:8 with a straightforward question and check it in a straightforward way: start the window on Monday, test on Tuesday morning, see what the number says. It is a reasonable instinct and it produces a lot of unnecessary discouragement, because the morning fasting reading is the one that responds least honestly to a longer overnight gap. Here is what the evidence actually shows, and which numbers are worth watching instead.

What the trials found

The most directly relevant study is a six-month randomised trial published in JAMA Network Open (opens in a new tab). Seventy-five adults with type 2 diabetes and obesity were assigned either to an 8-hour window — noon to 8 p.m., with no calorie counting at all — to daily calorie restriction, or to no change. HbA1c in the fasting group fell by 0.91% compared with the control group. The calorie-counting group fell by 0.94%. The two are, for practical purposes, the same result.

That equivalence is the finding, and it cuts both ways. 16:8 works — a 0.9% drop in HbA1c is a real, clinically meaningful change. It also did not beat ordinary calorie restriction on glucose. What it did beat was calorie counting on weight: the window group lost 3.56% of body weight relative to control, while the calorie-counting group's smaller loss did not reach statistical significance. The plausible reason is not metabolic magic but adherence — a clock is easier to keep to than a spreadsheet.

A smaller crossover trial in Diabetologia (opens in a new tab) adds a useful detail. Adults with type 2 diabetes who ate within a shorter, earlier daytime window for three weeks — last meal finished by 6 p.m. — spent about three more hours a day in the normal glucose range than when the same food was spread across fourteen hours or more. Their measured insulin sensitivity, though, did not significantly improve. The benefit appeared to come from when and how the eating was arranged, not from any change in how their bodies handled insulin.

So: yes, an eating window can lower blood sugar, by roughly as much as eating less would, mostly through weight and pattern rather than through anything the clock does on its own.

Why your morning reading may go up anyway

This is the part that catches people out. If your window ends at 8 p.m. and your first meal is at noon, your liver spends sixteen hours as your only source of glucose — and towards the end of that stretch it works harder, not less. The American Diabetes Association (opens in a new tab) describes what happens in the early hours: cortisol and growth hormone signal the liver to boost glucose production, an effect roughly half of all people with diabetes experience. A longer fast does not switch that off. It just means you meet it with an emptier stomach and often test right in the middle of it.

The ADA lists two other overnight causes worth knowing, because they lead somewhere different. Waning insulin — a long-acting dose that runs out before morning — produces a high reading that a dose or timing change fixes. And the Somogyi effect is a rebound: a low overnight prompts the body to release glucose in compensation, so the high morning number is a symptom of a hypo you slept through. That last one matters more once you are fasting sixteen hours, particularly if you take insulin or a sulfonylurea, and it is the reason a rising morning number is worth a conversation rather than a shrug.

The numbers that actually answer the question

Judge a window on the readings that reflect the whole day, not the single hardest hour of it. Four things are worth tracking, and only one of them is the morning number.

• Post-meal readings inside the window. Two hours after your largest meal. This is where a compressed window can quietly go wrong — three meals' worth of food inside eight hours can produce higher peaks than the same food spread out.
• The end of the fasting stretch. For the first two weeks, check before you break the fast, especially if you take insulin or a sulfonylurea. This is where lows show up if they are going to.
• Your morning fasting reading — as a trend. Watch the direction across a fortnight, not any single day. Expect it to be stubborn, and possibly to rise before it settles.
• HbA1c at three months. This is the honest scorecard. It averages every hour, including the fasting ones you never test, and it is the number the trials above were actually measuring.

Give it three months before deciding. Weight moves first, post-meal readings second, and HbA1c last — and if you judge the whole experiment on a Tuesday morning finger-prick in week one, you will almost certainly reach the wrong conclusion about something that was working.

Let three months of readings answer it

Log each reading with the time you first ate and last ate that day. The pattern that matters — post-meal peaks, end-of-fast lows, the direction of your morning number — only becomes visible across weeks.

Start logging with Diatic (opens Diatic on Google Play in a new tab)