GLP-1s During Ramadan: What the Four Studies Actually Tested

There is real evidence on taking a GLP‑1 through a month of dawn-to-dusk fasting, and it is narrower than it first looks. Four studies exist. All four enrolled people with type 2 diabetes. Not one enrolled somebody taking one of these drugs purely to lose weight — which describes most people reading this. What follows is what was actually measured, and where the gap between those people and you is doing the work.

The two randomized trials, and what they compared against

LIRA-Ramadan randomized adults with type 2 diabetes, at least ten weeks before Ramadan, either to switch to liraglutide 1.8 mg or to continue their sulfonylurea, both alongside metformin. Over Ramadan itself, short-term glucose control moved almost identically in the two groups: fructosamine fell 12.8 µmol/L on liraglutide and 16.4 on the sulfonylurea, a difference of 3.51 with a confidence interval running from −5.26 to 12.28 and a p-value of 0.43. Documented symptomatic low-blood-sugar episodes were reported by 2% of the liraglutide group — three people — against 11%, or eighteen people, on the sulfonylurea. Neither group had a single severe episode. Weight fell further on liraglutide, by about half a kilogram more.^[1]

The earlier Treat 4 Ramadan trial ran the same comparison across two UK sites in 99 adults. Its headline is quoted far more confidently than the trial supports: more people on liraglutide hit a composite of HbA1c under 7%, no weight gain and no severe hypoglycemia, but the odds ratio was 4.08 with a confidence interval of 0.97 to 17.22 and p = 0.06. That interval crosses 1. The result it did establish cleanly was self-recorded glucose at or below 3.9 mmol/L, which ran at roughly a third the rate on liraglutide, and again neither arm had a severe episode.^[2]

The problem is not the fast. It is the meal that ends it.

A 2026 study put continuous glucose monitors on people through Ramadan 2025 and found the damage was concentrated in one window. Fasting hours were not where control fell apart; the hours after iftar were. Among insulin-treated type 2 diabetes patients matched one-to-one, those also taking semaglutide or tirzepatide spent 74.4% of the time in range against 36.8% for insulin alone, and the four-hour glucose excursion after iftar was about 61% smaller. There was no increase in hypoglycemia and nobody stopped treatment.^[3]

The mechanism is not mysterious and it is the reason this finding travels beyond diabetes. These drugs slow gastric emptying and blunt the post-meal glucose rise. A fast that ends with a large, fast, carbohydrate-heavy meal is precisely the situation that mechanism is built for. It is also the situation that a month of restricted eating windows tends to produce.

If the drug is the oral one, the timing rule collides with the fast

Oral semaglutide carries an absorption protocol that a fasting day collides with. The tablet has to reach a stomach with nothing in it; the water it goes down with is capped at 120 mL; and for the next half hour nothing may follow it — not food, not a second drink, not another tablet. During Ramadan the only window that permits that is before dawn. A prospective study across the UAE, Saudi Arabia and Kuwait followed 257 people through it and found the protocol held up better than expected: of those keeping diaries, 68.4% followed the dosing instructions on at least 80% of days.^[4]

Adherence then showed up in the outcome, which is the practical point of the whole study. People who followed the instructions lost 3.2 kg; people who did not lost 1.6 kg. Their HbA1c fell 0.3 percentage points against 0.1, and the second of those was not statistically distinguishable from no change at all. Self-reported low-blood-sugar events were recorded in 31% of participants and none was severe — and the people reporting them were more likely to also be on a sulfonylurea or other glucose-lowering drugs.^[4]

What this does not tell you

It does not tell you what happens to somebody with no diabetes, on no other glucose-lowering medication, taking semaglutide or tirzepatide for weight, who stops drinking water for fourteen hours a day for a month. Every hypoglycemia figure above comes from people whose other medications carry that risk on their own, which is exactly what makes those figures reassuring and exactly what makes them inapplicable. And dehydration — the part of a dry fast that has nothing to do with glucose — is not an endpoint any of these four studies measured. Who is actually at risk of low blood sugar on one of these drugs, and who is not, is worked through in GLP-1 drugs and low blood sugar.

The practical questions worth taking to the person who prescribes for you are narrower and answerable: which of your medicines actually carry a hypoglycemia risk, whether an injection day should move, whether the pre-dawn window can hold an oral dose, and what your fluid intake looks like across two eating windows. Whether to fast is not a question this page is in a position to answer. If the month also involves travel across time zones, the injection-timing question is a separate one and is covered in travel and time zones on a GLP-1.