16:8 vs 5:2 Fasting: Which Works Better for Weight Loss?

Neither schedule has a proven long-term weight-loss advantage over the other, because the two have barely been tested against each other. The one randomized head-to-head trial we found lasted 4 weeks and enrolled 68 women: weight fell 2.7 kg across both groups, with reductions greater over time on 5:2, and the authors concluded overall effectiveness was similar ^[1]. Longer trials compare each schedule with something else, and both tend to land near ordinary calorie restriction ^[3][6].

The one direct trial

The Korean-registered trial by Kim and colleagues, published in Obesity in 2026, assigned women with overweight or obesity to 16:8 or to 5:2 for 4 weeks, without blinding ^[1]. Body weight fell 2.7 kg (95% CI -2.9 to -2.4), fat mass 1.7 kg and muscle mass 0.6 kg across both groups ^[1]. Weight and fat mass reductions were greater over time on 5:2, while muscle mass changes were similar. Waist circumference and blood pressure showed no between-group differences ^[1].

Three limits matter. Four weeks says little about what happens at a year. Everyone enrolled was a woman. And the 5:2 arm used a fasting-mimicking diet on fasting days ^[1], which may not match what someone doing 5:2 at home would eat. The abstract does not report a kilogram difference between the groups, so we cannot tell you how large the 5:2 edge was.

A second head-to-head study is on the way in type 2 diabetes: a published protocol describes 56 adults randomized to intermittent energy restriction or an 8-hour eating window, with feasibility and acceptability as the primary outcomes ^[17]. It has no results in that paper, and its fast-day pattern allows 2 to 3 days a week ^[17].

What each schedule did in its own trials

Everything else is indirect. The table sets the longer trials side by side, using each trial’s reported result ^{[2][3][4][5][6][7][8]}. The comparators differ, so the rows cannot be ranked against each other.

The range inside each schedule is wider than any gap between them. An 8-hour window produced under 1 kg in TREAT, where no calorie target was given ^[2], and 8.0 kg in Liu, where an earlier 8-hour window (8 am to 4 pm) was paired with a calorie target ^[4]. 5:2-style restriction produced 6.4 kg in Harvie, where both arms followed a 25% energy restriction ^[6], and 1.9 kg in Hajek, where people received one-off self-help instructions ^[8]. How much support and calorie structure came with the schedule looks at least as important as the schedule itself, though no trial here tested that directly.

Network meta-analyses do not settle it

Network meta-analyses try to compare schedules indirectly, through shared comparators. Their conclusions disagree on 16:8-style and 5:2-style fasting, and they agree mainly on something else: alternate-day fasting tends to rank highest.

• Huang 2024 (47 trials, 3,363 participants): alternate-day fasting had the largest mean difference (-3.42 kg), then time-restricted eating (-2.25 kg), short-term fasting (-1.87 kg) and continuous restriction (-1.59 kg) ^[9]. The abstract does not define short-term fasting as 5:2 specifically, so read this as a loose proxy.
• Elortegui Pascual 2023 (24 trials, 1,768 participants): an exploratory analysis ranked alternate-day fasting first, then continuous restriction, then time-restricted eating, with intermittent regimens overall similar to continuous restriction (mean difference 0.26 kg, 95% CI -0.31 to 0.84) ^[10]. The abstract’s ranking does not place 5:2.
• Ma 2024 (16 trials, 1,228 patients): 5:2 beat continuous restriction on BMI, alternate-day fasting beat 5:2 on waist circumference, and fat mass and dropout were comparable across forms. The authors rated study quality as low to moderate ^[11].
• Chen 2024 (umbrella review and network analysis, 153 original studies): all forms reduced weight compared with usual diets, and alternate-day fasting had the highest overall ranking for metabolic health ^[12].

Rankings like these describe probability of being best, and small differences can flip with the choice of trials. Our reading is that this literature does not support picking one of these two schedules for greater weight loss. The 2026 Cochrane review, whose headline estimate pools intermittent fasting forms, found little to no difference from regular dietary advice (mean difference -0.33% of baseline weight, 95% CI -0.92 to 0.26; low-certainty evidence) ^[13].

Lean mass

Both schedules appear to take some fat-free tissue along with fat. A 2025 meta-analysis of 20 trials put time-restricted eating at -0.58 kg of fat-free mass compared with no restriction ^[14], and TREAT found a significant between-group difference in appendicular lean mass index (-0.16 kg/m²) in its 50-person in-person cohort ^[2]. A meta-analysis of 5:2 trials also reported reduced fat-free mass, without a figure in its abstract ^[15]. Huang found a lean-mass decline with short-term fasting (-1.26 kg) ^[9]. The only direct comparison, Kim, found similar muscle changes in the two arms over 4 weeks ^[1]. That is too short and too small to say either schedule is kinder to muscle.

Sticking with it

Adherence may be the most practical way to choose, and the data here are thin. In Hajek, adherence to 5:2 self-help was 74% at 6 weeks, 31% at 6 months and 22% at one year; six weekly group sessions helped early but the weight difference was no longer significant at one year ^[8]. People still rated 5:2 instructions significantly higher than standard advice ^[8]. In Pavlou, participants kept to the 8-hour window on a mean of 6.1 days a week over 6 months, though that was a supported trial in a different population ^[5]. These figures come from different trials with different definitions of adherence, so they do not show that 16:8 is easier to keep up than 5:2.

Practical fit is where the schedules clearly differ, and a reader can judge this without a trial. 16:8 asks for the same routine every day and may suit people who skip breakfast anyway. 5:2 leaves five days alone but asks for two days of very low intake, such as the 500 to 600 kcal used in Carter ^[7]. Over 6 months or longer, intermittent fasting and continuous restriction were comparably effective for body weight and adiposity in a meta-analysis of 24 studies ^[16], so a plain daily calorie target remains a reasonable alternative to both.

Safety questions to raise with a prescriber

The clearest caution involves glucose-lowering medication. Carter reduced medications likely to cause hypoglycemia at baseline by protocol, and hypoglycemic or hyperglycemic events in the first 2 weeks still affected 35% of participants using sulfonylureas or insulin, with no significant difference between diets ^[7]. Anyone on those drugs should discuss dose changes before starting either schedule. A 5:2 meta-analysis reported no serious adverse events but mild physical and psychological side effects during fasting periods ^[15], and Cochrane rated the evidence on adverse events as very uncertain ^[13].

It also makes sense to ask before starting if you are pregnant or breastfeeding, have a history of an eating disorder, or take medication that must be taken with food. People on a GLP-1 face a separate set of tradeoffs around protein intake and appetite, which we cover in our review of stacking GLP-1 medication with fasting.