Intermittent Fasting vs Calorie Restriction: Trials

Trials mostly find that intermittent fasting and ordinary daily calorie restriction lead to similar weight loss. Of three head-to-head trials, two found similar results and one found modestly more loss with a fasting pattern. Reviews add that the evidence is short-term and often low certainty.

This is a general overview of published research, not medical advice, and the trials describe what researchers tested. They do not recommend a pattern to any reader. This site’s calculators never output a target below 1,500 kcal (men) or 1,200 kcal (women), and some fasting protocols in the research go well below those floors.

We read only the abstracts of the trials and reviews cited here, through PubMed. That limits how much detail we can report, and where a detail sits only in a full text or a press release, we left it out.

What “fasting diet” usually means, and what this article covers

“Fasting diet” is an umbrella phrase. In the studies below it covers three families of patterns.

• Time-restricted eating confines eating to a set window each day. The meal timing and circadian rhythm article covers windows and the terminology in detail, so we do not repeat that here.
• Alternate-day fasting alternates fast days with feast days. In one trial fast days meant a fraction of estimated energy needs, and feast days meant eating above them.
• Modified fasting on two or three days a week, such as the 4:3 and 5:2 patterns, restricts intake on a few days a week. In the 4:3 trial below, the other four days were unrestricted.

The comparison group is usually “continuous” or “daily” calorie restriction, which means eating a smaller amount every day. Some trials, such as Catenacci 2025 below, match the weekly calorie deficit between the two arms, so the question becomes whether the pattern of eating matters once the total is similar.

We did not review one-meal-a-day eating or fasts lasting several days, and this article makes no claim about them. Our article on meal frequency and metabolism covers the background on how often people eat. The calorie cycling article covers earlier meta-analyses of intermittent versus continuous restriction, so we focus here on individual trials and newer reviews.

What the 2026 Cochrane review does and does not show

The most recent large review is a Cochrane systematic review by Garegnani and colleagues, published in February 2026 (PMID 41692034). It included randomised trials of intermittent fasting in adults with overweight or obesity. The fasting types were time-restricted feeding, periodic fasting, alternate-day fasting and modified alternate-day fasting. Each intervention lasted at least four weeks, and each trial followed people for at least six months.

The review found 22 studies with 1,995 participants, published between 2016 and 2024. Outcomes were measured up to 12 months. Its main results, as the abstract reports them:

• Against regular dietary advice, fasting may result in little to no difference in percentage weight loss from baseline (mean difference -0.33, 95% CI -0.92 to 0.26; 21 studies, 1,430 participants). The certainty was low.
• Quality of life showed little to no difference against regular advice (low certainty).
• Adverse events showed little to no difference against regular advice, but the evidence was very uncertain (risk ratio 1.45, 95% CI 0.64 to 3.28; very low certainty).
• Against no intervention or a waiting list, the mean difference in percentage weight loss was -3.42 (95% CI -4.95 to -1.90; 6 studies, 427 participants), rated moderate certainty. The abstract’s own summary sentence for this comparison says “little to no difference”, which does not match its interval. We report the numbers and note the mismatch. The abstract does not spell out the sign convention.

The review did not compare intermittent fasting with continuous calorie restriction as a named comparator. Its comparator was “regular dietary advice”, and the abstract does not say what that advice contained. It also says none of the included studies reported participant satisfaction or diabetes status, and the abstract does not report adherence or lean mass. Its authors also noted that the studies covered short-term effects only and called for follow-up beyond 12 months.

The population matters too. These results apply to adults with overweight or obesity, and they should not be stretched to people of lower weight, older adults or teenagers.

Head-to-head trials against ordinary calorie restriction

Head-to-head trials give a more direct answer. Three randomised trials followed people for 12 months.

Trepanowski 2017. This single-centre trial (PMID 28459931) ran a 6-month weight-loss phase and a 6-month maintenance phase. On fast days the alternate-day group ate 25% of estimated energy needs, and on feast days 125%. The daily restriction group ate 75% of estimated needs every day. The authors concluded that alternate-day fasting did not produce superior adherence, weight loss, weight maintenance or cardioprotection. Dr Varady, one of the authors, disclosed an advance for a book about alternate-day fasting.

Catenacci 2025. This trial (PMID 40163873) asked the 4:3 group to restrict energy by 80% on three non-consecutive days a week, with unrestricted eating on the other four. The daily group cut intake by 34% to match the weekly deficit. Both groups received an intensive behavioural weight-loss programme and a recommendation to build up to 300 minutes of moderate activity a week. The abstract states “limited generalizability” as its limitation. Of 165 people randomised, 125 completed the trial.

Liu 2022. In this trial (PMID 35443107), everyone followed a calorie-restricted diet for 12 months: 1,500 to 1,800 kcal a day for men and 1,200 to 1,500 for women. One group also ate only between 8:00 a.m. and 4:00 p.m. Those bands match the floors this site uses, but they are trial parameters, not targets for a reader. The authors concluded that time-restricted eating was not more beneficial than daily calorie restriction for weight, body fat or metabolic risk factors.

Time-restricted eating without a calorie difference. Two shorter trials asked whether the eating window matters on its own. Maruthur 2024 (PMID 38639542) fed 41 adults with obesity and prediabetes or diet-controlled diabetes the same total calories for 12 weeks. Weight fell by 2.3 kg with a 10-hour window and 2.6 kg with a usual pattern, a difference of 0.3 kg (95% CI -1.2 to 1.9). Its authors suggested that earlier weight effects may be due to eating less. Lowe 2020 (PMID 32986097) randomised 116 adults to a 16:8 pattern or consistent meal timing for 12 weeks and found no significant difference in weight between groups (-0.26 kg, 95% CI -1.30 to 0.78).

Pooled analyses. A 2024 meta-analysis in Nutrients (PMID 39458528) pooled 10 trials with 623 adults with obesity. Both approaches led to 5.5 to 6.5 kg of weight loss at six months. Fasting-based strategies produced slightly greater short-term reductions in weight (-0.94 kg) and fat mass (-1.08 kg), which the authors called not clinically significant, and did not show superior long-term outcomes. A network meta-analysis by Huang and colleagues (PMID 39327619; 47 trials, 3,363 participants) estimated body-weight mean differences of -3.42 kg for alternate-day fasting, -2.25 for time-restricted eating, -1.87 for short-term fasting and -1.59 for continuous restriction. The abstract concluded that all four regimens were effective, and that regain occurred by four to six months in some subgroups. It did not report a direct pairwise comparison for each regimen, and it stated there is no consensus on an optimal one. That -3.42 is a different quantity from the Cochrane figure above.

The pattern across these sources is consistent in one respect: differences between fasting and daily restriction are small, and they do not all point the same way.

Is fasting easier to stick to?

Adherence is where the trials disagree most clearly. In Trepanowski 2017, 13 of 34 people in the alternate-day group dropped out (38%), against 10 of 35 in the daily restriction group (29%) and 8 of 31 controls (26%). The fasting group also ate more than prescribed on fast days and less than prescribed on feast days, while the daily group generally met its energy goals.

The Catenacci 2025 abstract reports only that 125 of 165 people completed the trial, without a split by group.

Harvard’s Nutrition Source page on intermittent fasting summarises a systematic review of intermittent fasting studies. That review found no significant difference in dropout between fasting groups and continuous restriction groups. The page states that fasting was not necessarily easier to follow than other weight-loss approaches. The calorie cycling article discusses adherence in the context of varying calories day to day.

Adherence can look different across studies, and a behavioural programme in both arms, as in Catenacci 2025, may change how easy either pattern feels.

What about lean mass?

Lean mass findings are mixed, and the two sides are worth naming.

Reports of lower fat-free or lean mass:

• Patikorn 2021 (PMID 34919135), an umbrella review of 11 meta-analyses covering 130 trials, found intermittent fasting associated with reduced fat-free mass. Median follow-up was three months.
• Huang 2024 found short-term fasting reduced lean mass (mean difference -1.26 kg, 95% CI -2.16 to -0.47).
• Xing 2026 (PMID 42280443), a review of 28 trials with 1,833 participants, reported fat-free mass loss in both younger and older cohorts. The certainty was low to very low, and the authors called the findings hypothesis-generating.
• In Lowe 2020, an in-person subset of 50 people had a lower appendicular lean mass index with time-restricted eating (-0.16 kg/m², 95% CI -0.27 to -0.05).

Reports of no significant difference:

• Liu 2022 found body lean mass results consistent with its main finding: no significant difference between groups.
• The 2024 Nutrients meta-analysis reported similar lean mass for fasting-based and continuous restriction.
• A 2026 meta-analysis of time-restricted eating combined with exercise (Nutrition Reviews, PMID 42743072; 11 trials, 602 participants) found no significant effect on fat-free mass.
• A 2026 meta-analysis of exercise combined with intermittent fasting (Frontiers in Nutrition, PMID 41883415; 65 trials, 3,293 participants) also found no significant effect on fat-free or lean body mass.

These studies differ in design, comparison group and duration, so they cannot simply be added together. A Harvard Health blog piece from 2022, written by a physician as opinion, says weight loss from any intervention often includes some loss of lean muscle mass and advises pairing fasting with resistance training. That is one physician’s view, not a trial result.

What the research cannot tell us yet

Several limits apply to nearly all of this evidence.

• Short follow-up. The Cochrane review followed people for up to 12 months, and the umbrella review’s median follow-up was three months. None of the abstracts reports long-term outcomes such as heart events.
• Certainty ratings. In the Patikorn 2021 umbrella review, 75 of 104 associations rested on very low-quality evidence and 22 on low-quality evidence, and only one was supported by high-quality evidence. The review’s authors still concluded that the beneficial associations supported intermittent fasting, especially modified alternate-day fasting, as a weight-loss approach for adults with overweight or obesity. Xing 2026 rated its certainty low to very low.
• Who was studied. Trepanowski 2017 was 86% women and Catenacci 2025 was 73.9% female. Both studies enrolled adults with overweight or obesity, so results may not carry over to other groups.
• Cholesterol. In Trepanowski 2017, LDL cholesterol was significantly higher in the alternate-day group at month 12 (11.5 mg/dL, 95% CI 1.9 to 21.1). Xing 2026 reported raised LDL cholesterol in older adults in a conservative sensitivity analysis, again with low-certainty evidence.
• Abstract-level reading. We read abstracts, not full papers, so the fine print on dropout, adverse events and adherence is beyond what we report.

We make no claims here about longevity, lifespan or effects beyond the outcomes listed above.

Who should not fast, or should talk to a clinician first

General information only, not medical advice. Anyone considering changing how they eat should talk to a doctor or registered dietitian. This section draws only on the pages we fetched for this article.

• Eating-disorder history. Know Diabetes and MyHealth London, two regional patient-information sites and not national NHS guidance pages, say people with a history of eating disorders such as anorexia and bulimia should not attempt intermittent fasting. The British Heart Foundation (BHF, page updated December 2025) lists a history of eating disorders among reasons no type of intermittent fasting is suitable. Harvard’s Nutrition Source, linked above, lists restrictive eating disorders too. A 2023 commentary in Clinical Diabetes and Endocrinology (Blumberg and colleagues, PMID 37865786, an opinion piece and not a trial) advised that people with an eating disorder history should never be encouraged to fast, and urged extreme caution for adolescents and young adults, particularly those who identify as female or gender diverse.
• Diabetes and glucose-lowering medicine. BHF says no type of intermittent fasting is suitable for people with diabetes or low blood sugar. Know Diabetes says a fasting plan could significantly increase the risk of hypoglycaemia for people on insulin, sulphonylureas or glinides, and that the diabetes team should be told about any diet change. We give no diabetes or medication advice here.
• Pregnancy and breastfeeding. Know Diabetes, MyHealth London, BHF and Harvard’s Nutrition Source all say to avoid fasting.
• Children and teenagers. BHF lists children and teenagers, and Harvard’s Nutrition Source lists adolescents in an active growth stage.
• Underweight. BHF lists being underweight. Harvard Health’s 2020 piece on older adults quotes a dietitian’s concern about weight loss in people already at a marginal body weight.
• Older adults. That same 2020 Harvard Health article says solid evidence about fasting in older adults is lacking, since human studies mostly involved younger or middle-aged people over short periods.
• Medicines that need food. BHF suggests checking with a doctor for anyone taking medicines with food. Harvard’s Nutrition Source and the 2020 Harvard Health piece both name medicines that need food, and the latter quotes a geriatrician’s concern about heart and blood-pressure drugs.

What we could not source. We did not review national guidance from NHS.uk, NIH, CDC, WHO, NICE or the American Diabetes Association, so nothing here reflects those bodies. Our older-adult and teenager statements rest on one or two sources each, and the pregnancy statement rests on regional, charity and university pages. We found no obstetric guidance.

How this fits with calorie targets on this site

Timing and total intake are separate questions. When trials matched the weekly deficit, as in Catenacci 2025, or matched calories exactly, as in Maruthur 2024, differences in weight were small. The calorie total behind a plan still does much of the work, and the calorie deficit calculator estimates a daily target from your own numbers. As weight changes, the guide to adjusting your calorie target explains how to update it.

The calculators clamp any result to 1,500 kcal for men and 1,200 kcal for women, and explain when they do. The percentages used in some fasting trials, such as 25% of energy needs on a fast day, would land below those floors for most adults. That is why this article reports the research and does not adapt it into a plan. For personal decisions about how to eat, a doctor or registered dietitian is the right person to ask.