Intermittent fasting: what the randomized trials found

Intermittent fasting has now been tested in enough randomized trials to give a fairly clear answer. In adults with overweight or obesity, it does produce weight loss, usually a few kilograms over several months. But when it is compared directly with ordinary daily calorie restriction, the two approaches lose about the same amount of weight, and changes in blood pressure, cholesterol, and blood sugar are similar as well. The trials have not shown that fasting does anything that eating less in a conventional way cannot do.

The evidence has real limits. Most trials are small, last three to twelve months, and recruit people with obesity rather than the general public. No randomized trial has followed people long enough to count heart attacks, strokes, or deaths. The fair summary is that intermittent fasting is one reasonable way to eat fewer calories, with modest and short-term evidence behind it, rather than a method with special metabolic effects.

This article is for adults who are considering a 16:8 schedule, alternate-day fasting, or a 5:2 pattern and want to know what the controlled studies actually measured before deciding.

What the trials mean by intermittent fasting

The term covers several patterns, and the trials test them separately. Time-restricted eating confines all food to a daily window, most often eight hours, with no rules about how much is eaten inside the window. Alternate-day fasting alternates a very low-calorie day, usually about a quarter of normal intake, with a day of normal or generous eating. The 5:2 pattern uses two low-calorie days a week. Trials of each are usually compared either with eating whenever one likes, or with daily calorie restriction, in which every day's intake is cut by a fixed amount.

Fasting compared with eating at any time

The clearest test of whether a time window alone causes weight loss is the TREAT trial, run largely through a smartphone app in the United States. Lowe et al., 2020 randomized 116 adults with overweight or obesity to eat only between noon and 8 p.m. or to eat three structured meals a day, for 12 weeks, with no calorie targets in either group. The fasting group lost 0.94 kg and the comparison group lost 0.68 kg, a difference of about a quarter of a kilogram that was not statistically significant. In a subgroup tested in person, the fasting group also lost a small amount of lean mass. The authors concluded that time-restricted eating on its own was not more effective than eating through the day.

A longer trial with a no-intervention control looked better for fasting. Lin et al., 2023 randomized 90 adults with obesity to an eight-hour eating window with no calorie counting, to daily calorie restriction of 25 percent, or to a control group eating over ten or more hours, for 12 months. Compared with the control group, the time-restricted group lost 4.6 kg and the calorie restriction group lost 5.4 kg. People in the fasting group had cut their intake by about 425 calories a day without being asked to, which is the most likely reason the window worked.

Fasting compared with daily calorie restriction

When fasting is compared with counting calories every day, the two methods have repeatedly come out level. The largest direct comparison is a 12-month trial in China. Liu et al., 2022 assigned 139 adults with obesity either to eat only between 8 a.m. and 4 p.m. while following a calorie-restricted diet, or to follow the same calorie restriction with no time limit. At one year the time-restricted group had lost 8.0 kg and the calorie-only group 6.3 kg. The 1.8 kg gap was within the range of chance, and waist size, body fat, blood pressure, and metabolic markers were also similar between the groups.

Alternate-day fasting has been tested the same way. Trepanowski et al., 2017 randomized 100 adults with obesity to alternate-day fasting, to daily calorie restriction, or to no intervention for a year. Weight loss relative to the control group was 6.0 percent with alternate-day fasting and 5.3 percent with daily restriction, a difference that was not significant. The fasting group had the highest dropout rate, at 38 percent, and tended to eat more than allowed on fast days and less than allowed on feast days. By 12 months their LDL cholesterol was 11.5 mg/dL higher than in the daily restriction group, which the authors flagged as a concern rather than a benefit. The Lin trial described above showed the same pattern, with the time-restricted and calorie restriction groups ending the year within a kilogram of each other.

Whether the timing of the window matters

One open question is whether eating earlier in the day works better than eating later. Jamshed et al., 2022 randomized 90 adults with obesity, all of whom received weight-loss counseling and a calorie target, to an early eight-hour window from 7 a.m. to 3 p.m. or to a window of 12 hours or more, for 14 weeks. The early-window group lost 2.3 kg more and had a 4 mm Hg larger fall in diastolic blood pressure. Body fat did not differ significantly, and the authors estimated the effect was equivalent to cutting about 214 extra calories a day. This is a single 14-week trial with a strict early window that many people would find hard to keep, so it is a lead rather than a settled result.

What the pooled reviews conclude

Several groups have combined the trials. An umbrella review of 11 meta-analyses covering 130 randomized trials, Patikorn et al., 2021, found that the median trial had only 38 participants and lasted about three months. Of 104 associations between fasting and health outcomes, only one was backed by high-quality evidence, a moderate drop in body mass index with modified alternate-day fasting over one to two months. Most of the rest rested on low or very low quality evidence, and fasting was also linked to a loss of fat-free mass.

A Cochrane review, 2021 on fasting for cardiovascular prevention pooled 18 trials with 1,125 participants. Fasting beat unrestricted eating by about 2.9 kg in the short term, but compared with continuous calorie restriction the difference was under 1 kg and judged very uncertain. No trial reported heart attacks, strokes, or deaths, and the reviewers did not consider the weight and glucose changes clinically significant. A newer Cochrane review, 2026 of 22 trials with 1,995 adults with overweight or obesity reached a similar conclusion. Compared with regular dietary advice, intermittent fasting may make little or no difference to percentage weight loss, with the evidence rated low certainty because of risk of bias in the trials.

Side effects and who should be careful

Reported side effects in the trials were mostly mild. The 2021 Cochrane review noted that of the four trials reporting side effects, the main complaint was mild headaches, and the 2026 review found adverse event rates that did not clearly differ from comparison groups, although it rated that evidence as very uncertain. Two findings deserve more attention. Several trials, including TREAT and the trials in the umbrella review, recorded a small loss of lean mass alongside fat, which matters most for older adults. And dropout was highest in the alternate-day fasting group of the one-year trial, which suggests that many people cannot sustain it.

People who take insulin or other glucose-lowering drugs, people who are pregnant or breastfeeding, anyone with a history of disordered eating, and people with conditions that make long gaps between meals risky were generally excluded from these trials and should talk to their doctor before trying any fasting pattern. For everyone else, the trials suggest that choosing between fasting and conventional calorie restriction is mostly a matter of which one is easier to keep up, since the measured results are about the same.

Sources

• Calorie Restriction with or without Time-Restricted Eating in Weight Loss (Liu et al., New England Journal of Medicine, 2022) — randomized trial, n=139 adults with obesity, 12 months, 8 a.m. to 4 p.m. window plus calorie restriction vs calorie restriction alone
• Effects of Time-Restricted Eating on Weight Loss and Other Metabolic Parameters: The TREAT Randomized Clinical Trial (Lowe et al., JAMA Internal Medicine, 2020) — randomized trial, n=116, 12 weeks, 16:8 window vs three meals a day, no calorie targets
• Effect of Alternate-Day Fasting on Weight Loss, Weight Maintenance, and Cardioprotection Among Metabolically Healthy Obese Adults (Trepanowski et al., JAMA Internal Medicine, 2017) — randomized trial, n=100, 12 months, alternate-day fasting vs daily calorie restriction vs control
• Time-Restricted Eating Without Calorie Counting for Weight Loss in a Racially Diverse Population (Lin et al., Annals of Internal Medicine, 2023) — randomized trial, n=90 adults with obesity, 12 months, 8-hour window vs 25 percent calorie restriction vs control
• Effectiveness of Early Time-Restricted Eating for Weight Loss, Fat Loss, and Cardiometabolic Health in Adults With Obesity (Jamshed et al., JAMA Internal Medicine, 2022) — randomized trial, n=90, 14 weeks, 7 a.m. to 3 p.m. window vs 12-hour or longer window, both with calorie restriction
• Intermittent Fasting and Obesity-Related Health Outcomes: An Umbrella Review of Meta-analyses of Randomized Clinical Trials (Patikorn et al., JAMA Network Open, 2021) — umbrella review of 11 meta-analyses covering 130 randomized trials, graded with GRADE
• Intermittent fasting for the prevention of cardiovascular disease (Allaf et al., Cochrane Database of Systematic Reviews, 2021) — Cochrane systematic review, 18 trials pooled, 1,125 participants, 4 weeks to 6 months
• Intermittent fasting for adults with overweight or obesity (Garegnani et al., Cochrane Database of Systematic Reviews, 2026) — Cochrane systematic review, 22 randomized trials, 1,995 participants, at least 6 months follow-up