Intermittent Fasting and Type 2 Diabetes: What the Research Shows and Safety Questions to Ask First

What “intermittent fasting” actually means

Search for intermittent fasting and you will find a dozen different plans that share one idea: some part of the day, week, or month is set aside with little or no food. It is not a single diet. Two people who both say they are “doing intermittent fasting” can be following schedules that have almost nothing in common, which is one reason the research is hard to summarize.

The patterns that come up most often fall into two families. The first is time-restricted eating, where you eat every day but only inside a daily window. The most common example is 16:8, a 16-hour fast with all meals inside an 8-hour window. Gentler versions like 14:10 or 12:12 are also used. The second family is intermittent energy restriction, where some days are normal and some days are very low in calories. The 5:2 plan and alternate-day fasting both belong here. One review of the topic describes both as 500 to 600 calories on the low days, either every other day or on two days of the week.

A third pattern, often called OMAD for “one meal a day,” compresses all eating into a single meal. It shows up in social media far more than in the trials, so there is very little formal research to point to for it, and I would treat it as the least studied pattern on this list.

Schedules are typical examples from how these plans are described, not recommendations. The 500 to 600 calorie figure for low days follows how one secondary review describes 5:2 and alternate-day plans.

What the research shows so far

I will describe what I was able to read, and be clear about what I could not. Several of the main medical journal sites and the PubMed Central archive blocked my tools while I was researching this page, so I did not read every paper in full. Where a number comes from a review that summarizes other studies, I say so, and I do not claim to have checked the underlying trials myself.

The broad picture. A 2025 narrative review in Frontiers in Nutrition on intermittent fasting in type 2 diabetes (Dyńka and colleagues) pulls together randomized trials, meta-analyses, and systematic reviews. It is a narrative review, not a new meta-analysis, and it does not report a single combined participant count. It reports that, across the studies it cites, intermittent fasting has been associated with better glucose control, improved insulin sensitivity, and weight loss, with adherence comparable to or better than continuous calorie restriction. The authors also list limits: study length, populations, and fasting protocols vary widely, some studies found no significant benefit, and certain groups, such as children, are not well studied. The review also discloses funding from a private foundation, with financial ties for two authors, which is worth knowing when weighing how strongly it is worded.

Some numbers it cites. As summarized from that review, and not independently checked: a 2025 meta-analysis of 1,101 adults with type 2 diabetes or prediabetes found about a 4.56 kg drop in weight, a 0.81 percentage point drop in HbA1c, and a 0.36 mmol/L drop in fasting glucose compared with controls. A 2024 randomized trial in people with obesity and type 2 diabetes compared 16:8 and 14:10 time-restricted eating with a control group and reported weight loss of about 4.0%, 3.2%, and 0.6%, with HbA1c drops of about 0.50, 0.53, and 0.20 percentage points. A separate 2024 trial comparing time-restricted eating with dietitian guidance reported HbA1c drops of 0.4 versus 0.3 percentage points and weight drops of 1.7 versus 1.2 kg. The difference between the groups in that last trial is small.

The same review cites a 2023 randomized trial in people with insulin-treated type 2 diabetes called INTERFAST-2, which reported an HbA1c fall of 7.3 mmol/mol (about 0.7 percentage points) at 12 weeks versus 0.1 in the control group. The review did not give the sample size in the part I read, and this is a group where the safety concerns below apply most strongly, so I would not take it as a sign that fasting is simple for anyone on insulin. It was a monitored study, not a do-it-yourself plan.

What was said at the 2025 ADA sessions. A clinician commentary published July 18, 2025 by HCPLive discussed a 6-month trial presented at the American Diabetes Association Scientific Sessions in June 2025. According to that commentary, both time-restricted eating and intermittent energy restriction improved HbA1c, glucose area under the curve, and body weight in people with type 2 diabetes. The clinician interviewed also said there is little data showing intermittent fasting works better than changing what patients eat or starting exercise, and that willingness and adherence matter most. I could only read the commentary, not the trial itself, so I do not have the number of participants or the exact effect sizes.

Numbers described as “summarized” come from a review or commentary, not from my own reading of each trial. Always look at the original paper before relying on a figure.

Why the results may not be only about fasting

When researchers compare intermittent fasting with a plain diet, a fair question is whether the fasting did the work or whether eating less did. A MedStar Health dietitian article makes this point directly: modest weight loss can improve insulin sensitivity and glucose control, and those changes may explain many of the benefits rather than the fasting itself. The 2025 narrative review says intermittent fasting results are comparable to consistent calorie reduction, and a secondary summary of the ADA Standards describes the weight loss as comparable to ordinary calorie reduction plans. In plain terms, if you eat less in total, many measures tend to move in the right direction, however you arrange the eating.

The worked example further down shows how quickly a fasting pattern becomes a calorie deficit on paper. That is not a criticism of fasting. It is a reason to ask what you are really trying to change: the clock, the calories, the carbohydrate, or the habit of late-night snacking.

The clinician commentary above adds another honest point. For many patients, the question is not “does fasting work” but “why this instead of something with more long-term data,” and whether the person can keep doing it. A plan you can follow for a year usually beats a stricter plan you quit in a month.

What the evidence mostly cannot tell you. Most of these trials are short, running from a few weeks to several months. Many enroll people who are motivated, supervised, and closely monitored. Few follow people for years, and I did not find strong long-term outcome data on heart events, kidney disease, or other complications of diabetes for any fasting pattern. Studies also often exclude the people at highest risk of low blood sugar. So the research can tell us that some people did well in a supervised trial, which is useful, but it does not promise the same result for you.

Safety comes first: low blood sugar, medicines, and who should not fast

This is the section to read even if you skip the rest. The biggest diabetes-specific concern with fasting is hypoglycemia, low blood sugar. The MedStar article puts it as the biggest concern, especially for people taking insulin or certain diabetes medicines. A 2017 commentary in the Cleveland Clinic Journal of Medicine on fasting in patients with diabetes sorts glucose-lowering drugs by their risk of causing lows. It describes several classes as having a low risk when they are not combined with insulin or a sulfonylurea, and singles out insulin and sulfonylureas as the ones that need careful planning around fasting. It also notes that some drugs can cause dehydration when fasting limits water, and that stopping or cutting medicines too far can cause high blood sugar and, rarely, a dangerous condition called diabetic ketoacidosis.

I am not going to say what to do about any medicine, and nobody should take that from this page. The general point is this: if you take insulin, a sulfonylurea, or any medicine that can lower blood sugar, the timing and dose of those medicines has to be worked out with your care team before you change when you eat. Do not adjust a medicine on your own because a fasting plan says so. The Cleveland Clinic commentary and the narrative review both stress monitoring and clinician supervision in this group.

Who is commonly told to avoid fasting. The MedStar article lists people who take insulin or medicines that can cause low blood sugar, people who often have low blood sugar, pregnant or breastfeeding people, people with a history of disordered eating or an eating disorder, and people with conditions that require regular meals or snacks. The 2025 narrative review adds caution or avoidance in people who are underweight, older adults at risk of malnutrition, people recovering from surgery, and some rare metabolic disorders. I did not find the kidney disease question settled in the sources I could read, so if you have kidney disease, treat it as a question for your kidney or primary care clinician, not an assumption in either direction. The same goes for any other condition you manage with medicines and meals together.

Warning signs. MedStar lists shakiness, sweating, dizziness, confusion, and feeling faint as symptoms that need immediate attention. Your own care team will have told you what your low blood sugar feels like and what to do when it happens. If you do not have that plan yet, getting one is the first step before fasting, not after.

Ramadan and religious fasting. Many people with diabetes fast for religious reasons. The 2017 Cleveland Clinic commentary was written for that situation, and the 2025 narrative review cites a 2023 meta-analysis of Ramadan fasting in type 2 diabetes that reported an HbA1c drop of about 0.55 percentage points and a fasting glucose drop of about 12 mg/dL. A secondary article on the ADA’s 2025 Standards says they were updated to give more structured guidance on fasting, including a risk assessment before fasting and adjustments to medicines and glucose monitoring. I could not read the Standards themselves, so I cannot quote them. If you plan to fast for religious reasons, ask your clinician for a pre-fasting risk assessment and a written plan, including when you should break a fast.

Worked examples: what the schedules look like in numbers

These examples are illustrations of arithmetic, not targets or recommendations. They use round numbers so you can see how a schedule changes the clock and the totals. Your own needs depend on your size, activity, medicines, and health conditions, which only your care team can weigh.

Example 1: the eating window. A window is just the hours between the first bite and the last. If the window opens at the time shown, here is when it closes and how long the fast runs.

Computed from the opening time and window length. Real life moves around: a 16:8 window starting at 8:00 AM closes at 4:00 PM, which suits some schedules better than noon to 8:00 PM.

Example 2: the same day, split two ways. Suppose a person eats 1,800 calories and 135 grams of carbohydrate in a day, and compare spreading it across three meals with squeezing it into two. Three meals work out to about 600 calories and 45 grams of carbohydrate each. Two meals work out to about 900 calories and 68 grams each. The totals are identical, but each meal is about 50 percent bigger, which is why people who change their meal count often need to rethink portions and the balance of protein, fiber, and carbohydrate in each meal. The post-meal glucose tracker is one way to look at how bigger meals feel for you, if you test.

Example 3: why a fasting plan can mean a calorie deficit. Say a person eats about 2,000 calories on a normal day. On a 5:2 plan with 500 calories on two days, the week adds up to 11,000 calories, which averages about 1,571 calories a day, roughly 21 percent less than 2,000. On an alternate-day plan with 600 calories on every other day, a two-week stretch averages about 1,300 calories a day, roughly 35 percent less. A time-restricted plan, in contrast, only changes calories if the shorter window leads you to eat less, so it can come out anywhere from no change to a sizable drop. That is the reason weight and glucose results across studies look so different, and the reason it is hard to credit the clock alone.

Sample days and the clock

If you and your care team decide that a time-restricted pattern makes sense, here are some sample day shapes to talk through. They are examples of timing only. They say nothing about what or how much to eat, which depends on you.

Meal times are examples of the clock only, not eating advice.

If you wear a CGM: what you might see during a fast

A continuous glucose monitor can be useful during any change in eating pattern, and the 2025 narrative review recommends monitoring and clinician supervision for people on medicines that can cause lows. The guide to reading CGM data covers the basics of reading a trace. The table below describes situations that people commonly ask about. It is a set of reasonable questions to explore with your clinician, not a set of rules, and I have not verified how these play out for any one person.

This is a list of questions to explore, not a statement of what always happens.

A common question is what “breaks” a fast. There is no single agreed answer in the sources I read. Some people care about calories, some about insulin response, and some just about sticking to a rhythm. If you use a sweetener or a calorie-containing coffee drink, the sweeteners guide covers what is known about glucose and what is not. Whatever you choose, if you take medicines tied to meals, the definition of “fasting” should be one your care team agrees with.

Different starting points

If you have prediabetes

The studies cited above included people with type 2 diabetes or prediabetes, and a 2023 trial summarized in the narrative review reported improved glucose tolerance in 209 at-risk people that was sustained at 18 months with fasting combined with early time-restricted eating. That is promising, and it is one trial. Prediabetes also responds to many other lifestyle changes, and the diabetes food guide and the walking after meals guide cover options with their own evidence. If you are not on glucose-lowering medicines, your main safety questions are about eating disorders, pregnancy, and other health conditions, and your main practical question is whether you can keep it going.

If you have type 2 diabetes and take no glucose-lowering medicine

Low blood sugar from fasting is less of a worry when you are not on a drug that can cause it, but that is a question to confirm with your clinician, not to assume. This group is where most of the milder time-restricted studies live, with HbA1c changes in the range of a few tenths of a percentage point as summarized above. Some people also see practical benefits such as less night eating, which the MedStar dietitian notes. Others find fasting makes them overeat later, which the same article warns about.

If you take insulin or a sulfonylurea, or any medicine that can lower blood sugar

This is the clinician-first group. Do not start a fasting plan without a conversation about your medicines, your monitoring, and what to do if you go low or high. The research that exists for this group, such as INTERFAST-2 as summarized above, happened under supervision. If your care team says no, that is a reasonable answer, and there are other food and activity changes to work on. If they say yes, ask for the plan in writing.

If you use a CGM

A CGM can show whether your glucose moves the way you expect, and can alert you to lows. The 2025 narrative review recommends it for people who fast on medicines with a risk of lows. Remember that a sensor shows an estimate, not a diagnosis, and that overnight sensor errors happen. Read the CGM reading guide first so you know what a normal trace looks like, and keep a plan for unexpected readings.

Common mistakes

• Changing medicines without telling your team. This is the most serious one. A fasting plan does not replace a care plan.
• Making the fast the goal. The fast is a tool. If the real goal is steadier glucose, fewer late-night snacks, or weight loss, check whether the plan is moving that.
• Breaking the fast with a very large, high-carbohydrate meal. The MedStar article advises avoiding very large meals after fasting, and notes that some people overeat later in the day.
• Skipping protein, fiber, and vegetables. Fewer meals means each one carries more weight. The type 2 diabetes diet guide covers the plate in more detail.
• Fasting through symptoms. Shakiness, sweating, confusion, or feeling faint are signs to act on, not to push through.
• Copying a social media schedule. The loudest schedules, such as one meal a day, are the least studied.
• Assuming it works because it is trending. The research is real but modest, and it is mostly short term.
• Using it when you have a history of disordered eating. Sources list this as a reason to avoid fasting. Please talk to a professional first.

Myth versus fact

Questions to ask your care team before you start

1. Is intermittent fasting appropriate for me, given my health history and medicines?
2. Do any of my medicines raise the risk of low blood sugar if I skip or delay meals, and how should that be handled?
3. What should my glucose targets be while I fast, and when should I break a fast?
4. How should I test or monitor, and how often, during the first few weeks?
5. What low blood sugar symptoms should I look for, and what is my plan if I get them?
6. Are there health conditions I have, such as kidney disease, that change the answer?
7. Would a registered dietitian review my meals so each one still has enough protein, fiber, and nutrients?
8. How will we decide after a few weeks whether it is helping, and when should we stop?
9. If I want to fast for religious reasons, can we do a pre-fasting risk assessment and write a plan?
10. What are the options if fasting is not right for me?

The complete guide to type 2 diabetes pulls together the other pieces of care that this page does not cover.

The bottom line

Intermittent fasting is a legitimate research topic, and some adults with type 2 diabetes have done better on it in studies, with modest changes in HbA1c and weight. The evidence is short term and uneven, the effects look similar to ordinary calorie reduction in several sources, and the main safety risk, low blood sugar for people on certain medicines, is serious enough that the plan should come from your care team. If fasting appeals to you, treat it as one optional tool among several, start with a conversation, and choose a schedule you can keep. If it does not appeal to you, that is fine, because what you eat and how active you are still matter a great deal.

Frequently asked questions

Does intermittent fasting help with type 2 diabetes?

In some studies, versions of it have been followed by modest drops in HbA1c and weight in adults with type 2 diabetes. The evidence is mostly short term, the studies differ a lot, and several sources describe results as similar to ordinary calorie reduction. It is not a cure.

Is intermittent fasting safe if I take diabetes medicine?

It depends on the medicine. Insulin and sulfonylureas, and other medicines that can lower blood sugar, make low blood sugar a real risk when meals are skipped. Medicine timing and doses should be planned with your care team before you fast. Do not change medicines on your own.

Who should not do intermittent fasting?

Sources list people who take insulin or medicines that can cause low blood sugar, people who often have lows, pregnant or breastfeeding people, people with a history of disordered eating, and people whose conditions need regular meals. Other groups, such as underweight people and older adults at risk of malnutrition, are also told to be careful. Ask your clinician about your own situation.

What are the signs of low blood sugar while fasting?

Shakiness, sweating, dizziness, confusion, and feeling faint are listed as symptoms that need immediate attention. Your care team should give you a plan for what to do if you have them.

Which intermittent fasting schedule has the most research?

Time-restricted eating, such as 16:8 and 14:10, and intermittent energy restriction, such as 5:2 and alternate-day fasting, have the most published trials. One-meal-a-day plans have very little formal research.

Is 16:8 better than 14:10?

One 2024 trial summarized in a review reported similar HbA1c drops in the two groups, about 0.50 and 0.53 percentage points. That is one trial, so I would not say one is better. A schedule you can keep matters more.

Does fasting work better than just eating less?

The sources I read describe results as comparable to ordinary calorie reduction, and a clinician commentary says there is little data that intermittent fasting beats changing what you eat or exercising. Some people simply find a schedule easier to follow.

Can I use a CGM while fasting?

Yes, many people do, and a review recommends monitoring for people on medicines with a risk of lows. A CGM shows estimates, so confirm with a fingerstick if a reading seems off or you feel unwell, and bring unusual traces to your care team.

What about fasting for Ramadan or other religious reasons?

A 2017 Cleveland Clinic Journal of Medicine commentary addresses medicine and monitoring planning for patients who fast, and a secondary article says the ADA’s 2025 Standards added more structured fasting guidance. Ask your clinician for a pre-fasting risk assessment and a written plan.

How long should I try fasting before deciding if it works?

There is no standard answer in the sources I read, and the studies ran from weeks to several months. A reasonable approach is to agree on what you are measuring with your care team, check in after a few weeks, and stop if you have lows, symptoms, or trouble keeping up.

Sources checked

• Dyńka et al., Intermittent fasting in the treatment of type 2 diabetes, Frontiers in Nutrition, 2025
• Olansky, Strategies for management of intermittent fasting in patients with diabetes, Cleveland Clinic Journal of Medicine
• MedStar Health, Intermittent fasting and diabetes
• HCPLive, What clinicians need to know about intermittent fasting in type 2 diabetes (July 18, 2025)
• DiabetesEd, secondary summary of fasting guidance in ADA Standards of Care
• Effect of Intermittent Fasting on Glycaemic Control in Patients With Type 2 Diabetes Mellitus: systematic review and meta-analysis (listed; full text not read)

About the author: Charles Kirkland is the founder of Metabolic Clarity Labs and author of The Journey Back. He is not a physician. This is general education, not a medically reviewed treatment plan.

Sources checked October 31, 2026. Discuss personal decisions with a qualified clinician. Do not start, stop, or change prescribed treatment based on this article.