Low-carb vs low-fat: what the head-to-head trials settled

When trials randomly split people into a low-carb group or a low-fat group, weight loss came out close either way. In the weight-loss trials among 53 studies, low-carb dieters lost about one kilogram more than low-fat dieters. That is roughly two and a half pounds, not the gulf either camp claims.²

What the trials actually did

The strongest of these trials share one feature: nobody chose a side. People were assigned by coin-flip logic to a lower-carbohydrate plan or a lower-fat plan, then followed for months or years. Some let people eat until full, on purpose, to see what each plan did to hunger on its own. Others matched calories so the only real difference left was what filled the plate.

That design answers a narrower question than which diet works. It asks what happens when two people who would otherwise eat the same way are handed different instructions. Across the biggest of these comparisons, the answer is not much that a bathroom scale would notice.

The gap between camps

One review pooling 53 studies and more than 68,000 people found low-carbohydrate diets beat low-fat diets by 1.15 kg (range 0.52 to 1.79 kg) across 18 comparisons in weight-loss trials.² That is the headline number for the entire argument, and it is small.

A separate comparison of 48 trials and 7,286 people tracked named programs against no diet at all. Low-carb diets produced 8.73 kg of loss at six months and low-fat diets 7.99 kg, a modest gap with overlapping ranges.³ By twelve months the two had converged almost exactly, 7.25 kg against 7.27 kg.³ Whatever early edge low-carb held had gone by twelve months.

Weight loss on named diet programs, by follow-upkg lost

Low-carb, 6 months8.73 kg

Low-fat, 6 months7.99 kg

Low-carb, 12 months7.25 kg

Low-fat, 12 months7.27 kg

Where fat and carbs actually diverge

The two diets are not identical in their effects, just close on the scale. A large analysis of 174 trials found that cutting carbohydrate lowered triglycerides by about 15 mg/dL, and trimmed both systolic and diastolic blood pressure, while raising HDL cholesterol.¹ Ketogenic versions of the diet produced the most weight loss of any carbohydrate-restricted approach but also pushed LDL cholesterol up the most, a trade the moderate versions did not make as sharply.¹

None of that shows up on a bathroom scale. It shows up on a blood panel, and it is the part of the argument that "which diet loses more weight" tends to skip past entirely.

Blood sugar tips the scale

The exception where a clear winner emerged was in people already living with type 2 diabetes. Over six months, a low-carbohydrate, higher-fat diet improved blood sugar control by 6.1 mmol/mol more than a low-fat diet on the standard test.⁴ It also produced 3.8 kg more weight loss (range 1.4 to 6.2 kg), with no calorie counting required on either side.⁴

Extra weight lost on low-carb vs low-fat, kg0 = no effect

Weight-loss trials, 18 comparisons1.15 (0.52 to 1.79)

Type 2 diabetes, six months3.8 (1.4 to 6.2)

That advantage did not last on its own. Three months after the structured trial ended, the gains between groups were no longer holding.⁴ The diet worked while it was actively followed, which is a different claim than the diet having changed something permanent.

What drives appetite

The most useful small trials were not the ones ranking diets. They looked at why people eat more or less on each one.

One tightly controlled comparison put people through both diets, two weeks each, and let them eat exactly as much as they wanted. The 20 adults lived in a hospital unit for the month. Counter to what the popular theory about carbohydrate and insulin predicts, the low-fat, plant-based diet won out. People ate about 689 fewer calories a day on it than on the low-carbohydrate, animal-based diet over the first two weeks, and 544 fewer in the final week.⁵ In this trial, the idea that low-carb eating curbs appetite did not hold up.

A second trial, in 55 women with obesity and lipedema, measured appetite hormones directly. Both a low-carbohydrate and a low-fat version of the same low-calorie diet produced weight loss over eight weeks. The low-carbohydrate group lost more, 10.3% of body weight against 7.3%, alongside a drop in the hunger hormone ghrelin after meals that the low-fat group did not show.⁷ The two trials point in opposite directions on which diet quiets hunger more.

When only the macronutrient split differs, the diet barely matters

The plainest evidence against either side having a real edge comes from trials that compared different macronutrient splits. Across five such trials and ten study groups in people with type 2 diabetes, differing amounts of carbohydrate, protein and fat produced no meaningful difference in weight loss, blood sugar, cholesterol or blood pressure.⁶ The split was not what moved the numbers.

What remains unknown

Most of these trials ran for months, not years, and the one that measured a genuine advantage for low-carb eating in diabetes saw it start to fade within three months of the diet ending.⁴ Nothing here says what either diet does to a person who keeps it up for a decade.

The appetite findings are the most interesting and the least settled. Two small trials looked at appetite and pointed in opposite directions: less eating on low-fat in one, more weight lost on low-carb in the other.^5,7 Neither trial was large enough to say which people respond which way.

And the metabolic split — lower triglycerides and blood pressure against higher LDL on the more carbohydrate-restricted end — means "better" depends on which number a person is trying to move.¹ The trials settled the weight-loss argument as roughly a draw. They did not settle what to do with a body that responds differently on the inside than it does on the scale.