Does Keto Beat Other Diets at Equal Calories? What a 2026 Meta-Analysis Found

Keto may beat higher-carbohydrate diets at equal prescribed calories, but only slightly: a 2026 meta-analysis of six randomized trials and 259 participants found 1.49 kg more weight loss (95% CI −2.41 to −0.58).^[1] The authors rated the certainty of that evidence as low. Most trials were short, actual calorie and protein intake was uncertain, and the review did not establish that the extra weight lost was fat.^[1]

What the meta-analysis tested

Many keto comparisons let one group eat freely while the other counts calories, which muddies the result. Żurawski and Bartosiewicz kept only trials that assigned both groups the same energy intake or the same targeted deficit. Ketogenic diets had to provide no more than 50 g of carbohydrate a day or 10% of energy, and they were compared with higher-carbohydrate diets in adults with overweight or obesity.^[1] The search identified seven eligible studies, and six supplied data that could be pooled.

The results pointed the same way across trials, with no detectable heterogeneity (I² = 0%), and the estimate held when the trial at high risk of bias was removed. The authors caution that 0% does not mean the trials were clinically alike. None was at low risk of bias across all domains, and several analyses used only participants who finished.^[1] One point on definitions: the Golay trial qualified on grams of carbohydrate, about 37 g a day, even though that was 15% of its energy.^[1] Two studies did not report any biochemical check for ketosis, and a third did not clearly describe how ketosis was confirmed, so “ketogenic” here describes the diet as prescribed more reliably than the physiology participants reached.^[1]

Why a 1.49 kg edge does not prove a metabolic advantage

Equal prescriptions are different from equal intake. Only two trials provided all the food, and the others relied at least partly on free-living participants and mostly self-reported intake.^[1] A diet that reduces appetite could lead people to eat less than their counterparts, which would be a real practical benefit but would say nothing special about how the body handles a calorie.

Protein is the second problem. Only three of the six pooled trials met the review’s protein-matching criterion, and in the others the keto arm got a larger share of energy from protein. In those three matched trials, the difference was −1.79 kg, with a confidence interval of −5.69 to 2.11 that includes no difference at all.^[1] The authors conclude that this analysis neither confirms nor excludes an independent effect of cutting carbohydrate. They also note that early weight loss on very-low-carbohydrate diets may partly reflect glycogen and its associated water, and body composition was not pooled.^[1]

What tightly controlled feeding studies add

Kevin Hall’s metabolic-ward study is the closest look at the mechanism, though it is small and not randomized. Seventeen overweight or obese men ate a high-carbohydrate baseline diet for 4 weeks, then an isocaloric ketogenic diet with protein held steady for 4 weeks. Energy expenditure rose by 57 kcal a day measured in a chamber and 151 kcal a day by doubly labeled water, but body fat loss slowed on keto, and fat-free mass fell.^[2] The authors called those expenditure increases relatively small and near the limits of detection.

A broader analysis from Hall and Guo pooled 32 controlled feeding studies that swapped carbohydrate for fat at equal calories. Energy expenditure was 26 kcal a day greater and fat loss 16 g a day greater on the lower-fat diets, the opposite of what the carbohydrate-insulin argument predicts, though both differences are small.^[3] Hall disclosed funding from the Nutrition Science Initiative to study ketogenic diets and energy expenditure.^[3] Taken together with the protein-matched result above, these studies make a large calorie-independent fat-burning advantage hard to support. For the broader math of a calorie target, see our review of 1,500 calories a day.

What broader comparisons show

Other reviews do not require matched calorie prescriptions, and their numbers vary with their definitions and comparators:

• At 12 months or more: a 2013 meta-analysis of 13 trials found very-low-carbohydrate ketogenic diets beat low-fat diets by 0.91 kg (95% CI −1.65 to −0.17; 1,415 patients).^[4]
• Keto against control diets: a pooled analysis of 18 trials reported −2.87 kg (95% CI −3.84 to −1.89), with long-term effectiveness unclear.^[5]
• Keto against Mediterranean: a 2026 meta-analysis reported −3.56 kg (95% CI −5.62 to −1.49) favoring keto, and described the gain as possibly short-term.^[6] Its abstract does not say whether calories were matched. Our Mediterranean diet review covers that pattern’s cardiovascular record, which is its main strength.
• Carbohydrate dose-response: across 110 trials, each 10% cut in carbohydrate intake reduced weight by 0.64 kg at the 6-month follow-up and 1.15 kg at the 12-month follow-up. Beyond 12 months the relationship was non-linear, and intakes higher than 40% or lower than 30% of energy were not effective.^[7]

A 174-trial meta-analysis in 2025 reached a similar balance, finding that ketogenic diets produced greater weight loss than other carbohydrate-restricted patterns but also larger increases in LDL and total cholesterol.^[8]

The tradeoffs: LDL cholesterol and lean mass

Two 2026 meta-analyses of randomized trials landed on nearly the same LDL figure: a rise of 8.49 mg/dL across 62 studies^[9] and 8.22 mg/dL across 53 trials, the latter rated low certainty under GRADE.^[10] The same analyses found triglycerides falling by 19.96 and 22.31 mg/dL and HDL rising by 3.61 and 3.52 mg/dL.^[9][10] Neither has cardiovascular event data to show what the net effect means over years. A 2023 umbrella review rated the LDL increase as supported by high-quality evidence and clinically meaningful, while noting a median trial follow-up of only 13 weeks.^[11]

Lean mass findings are more mixed. One pooled analysis found fat-free mass down 0.81 kg and lean body mass down 0.63 kg, although fat mass fell too.^[5] The umbrella review found reduced muscle mass with ketogenic low-carbohydrate high-fat diets in healthy participants, but no worsening of muscle mass with very-low-calorie ketogenic diets in adults with overweight or obesity.^[11] An 8-week secondary analysis in adults 60 and older reported that intermuscular fat fell 24.4% on a weight-maintaining very-low-carbohydrate diet, versus 1.0% on a low-fat diet.^[12] That is a signal worth testing, not evidence about weight loss.

Keto while taking a GLP-1 drug

Very little controlled evidence addresses this combination. A 2026 scoping review of diet alongside semaglutide or tirzepatide found 12 studies and concluded that evidence on the best dietary strategy is sparse, while pointing to high-protein, nutrient-dense eating and early dietitian input.^[13] The most direct paper we found is a retrospective case series of seven adults on a ketogenic diet plus low-dose semaglutide, with mean loss of 21.9 kg and lean mass estimated by bioimpedance. It had no control group, its authors call the findings hypothesis-generating, and one author advises a ketone company.^[14] It cannot show that keto added anything to the drug. Whatever the carbohydrate level, it helps to have a daily protein target when appetite is low.

What to raise with a clinician

• Ask whether a lipid panel before starting and during the diet makes sense, since the LDL rise is the most consistent downside in the trial data.^[9][11]
• If you take a glucose-lowering medicine, ask whether doses or the plan should change first. The SGLT2 inhibitor case above is one reason.^[15]
• Ask how to protect muscle, particularly if you are older or on a GLP-1 drug, since lean mass findings are mixed.^[5][13]
• Treat the choice as one of fit. A difference of about 1.5 kg at matched calorie prescriptions, from low-certainty evidence, may matter less than which pattern you can sustain.^[1]