Fasting arrived with a naming system. 16:8, 5:2, alternate-day, one meal a day, extended. The names suggest a family of distinct interventions, each with its own mechanism. The trial evidence tells a simpler story. Start with what each one is. Time-restricted eating confines all food to a daily window, most commonly eight hours, with no calorie target attached. 5:2 means five ordinary days and two days of substantial restriction. Alternate-day fasting cycles a low-intake day against a normal day. One meal a day compresses everything into a single sitting. Extended fasting runs past 24 hours and often for several days, which puts it in a different category from the rest. Now the trials. The largest and longest test of time-restricted eating randomized 139 adults with obesity in Guangzhou for 12 months. One group ate inside an eight-hour window, from 8:00 to 16:00, alongside calorie restriction. The other group did calorie restriction alone. The window group lost 8.0 kg. The calorie-restriction group lost 6.3 kg. The difference between them was not statistically significant (Liu et al., New England Journal of Medicine, 2022). The TREAT trial reached the same place by a different route. 116 adults with overweight or obesity were assigned either a noon-to-eight eating window or three structured meals, for 12 weeks. The window produced a small weight difference that did not reach significance, and the food logs suggested it worked mainly by trimming intake somewhat (Lowe et al., JAMA Internal Medicine, 2020). Alternate-day fasting has the cleanest head-to-head comparison. 100 metabolically healthy adults with obesity were randomized for a full year across three arms: alternate-day fasting, daily calorie restriction, and no intervention. Alternate-day fasting did not beat daily calorie restriction on weight loss, on weight maintenance, or on cardiovascular risk markers. It also had a higher dropout rate (Trepanowski et al., JAMA Internal Medicine, 2017). 5:2 was tested in 107 overweight and obese premenopausal women over six months, comparing a 25% energy restriction delivered as two restricted days per week against the same restriction spread across all seven days. The two approaches produced comparable weight loss (Harvie et al., International Journal of Obesity, 2011). That trial is worth noting for a second reason: it was women-only, which is uncommon in this literature. Put those four together and a pattern emerges. These styles lower intake, and the intake is what moves the outcome. The clock is a delivery mechanism. Which reframes the choosing question. If the styles are roughly equivalent in effect, the useful variable is whether you can keep one. There, the data are less flattering. A pragmatic trial of 300 adults with obesity found that adherence to 5:2 ran at 74% after six weeks and 22% after one year, with weight loss of 1.9 kg versus 1.8 kg for standard advice (Hajek et al., PLOS One, 2021). Whatever the protocol promises, most people are no longer doing it a year later. One meal a day deserves a separate note, because it is the least studied. The main controlled work is an eight-week randomized crossover in healthy normal-weight middle-aged adults comparing three meals a day against one. Eating once a day significantly increased hunger, reduced fat mass, and raised blood pressure along with total, LDL and HDL cholesterol, while lowering cortisol (Stote et al., American Journal of Clinical Nutrition, 2007). That is a small crossover, and long-term outcomes for this pattern are essentially unstudied. Extended fasting sits apart. In an observational study of 1,422 people fasting between 4 and 21 days, adverse effects were reported in under 1% of participants (Wilhelmi de Toledo et al., PLOS One, 2019). That safety record arrives with daily medical supervision and a protocol supplying 200 to 250 kcal per day, so it should be read as a description of supervised fasting rather than a general assurance. A closing note on all of this. These are averages from groups of people who are not you, and several of the samples are small. Research describes central tendencies; you are the one running the experiment on yourself. Take the pattern that survives an ordinary week, and notice whether you are choosing it or obeying it.
Safety
Educational only, not medical advice. Talk to your doctor before changing how you eat, particularly if you take medication affecting blood sugar or blood pressure, are pregnant or breastfeeding, are underweight, or have a history of disordered eating. If you are struggling with your relationship to food, the National Alliance for Eating Disorders helpline can help.
References
Liu D et al., N Engl J Med, 2022 (PMID 35443107). Lowe DA et al., JAMA Intern Med, 2020 (PMID 32986097). Trepanowski JF et al., JAMA Intern Med, 2017 (PMID 28459931). Harvie MN et al., Int J Obes, 2011 (PMID 20921964). Hajek P et al., PLoS One, 2021;16(11):e0258853. Stote KS et al., Am J Clin Nutr, 2007;85(4):981-988. Wilhelmi de Toledo F et al., PLoS One, 2019 (PMID 30601864).