High-protein diet types and what their weight-loss results mean

High-protein diets differ in carbs, fat and exercise. Read the research on weight loss, lean mass and the challenge of keeping a diet going.

日本語で読む（Japanese version）

A high-protein, low-fat dinner and a high-protein, high-fat dinner can both be called high-protein. Yet their instructions about fat point in opposite directions. Looking only at the protein portion puts them on the same shelf; looking at the rice, pasta and cooking oil reveals different meals. Choosing dinner means deciding what to reduce as well as what to increase.[1][3]

The research also covers more than one kind of high-protein diet. Some trials change the balance of protein and carbohydrate within a low-fat diet. Others compare different fat levels or combine the diet with demanding exercise. The shared word “protein” does not make their weight-loss results interchangeable.[1][3][19]

The useful question is what changed in the comparison, and how much difference that change made. The feature on protein, fullness and muscle maintenance explores appetite and metabolism in greater detail. Here, the focus is how to compare diets with similar names, including the conditions that make a study relevant to an ordinary meal after work in Tokyo.

Start with what the diet replaces

Diet studies compare the proportions of protein, carbohydrate and fat, often called macronutrients. A higher protein proportion leaves several possibilities for the rest of the meal. Whether the whole diet also reduces energy intake is another condition. A comparison between energy-restricted, low-fat diets answers a different question from an analysis bringing together diets with varied carbohydrate and fat levels.[1][3]

Imagine increasing the main dish at dinner while reducing the rice. That is a different change from increasing the main dish and keeping the rice unchanged. If researchers prescribe the same total energy to both groups, their findings cannot support adding unlimited protein to the food someone already eats. The meta-analysis of high-protein, low-fat diets compared diets prescribed with equal energy and matched fat intake.[1]

To compare the protein-to-carbohydrate balance in low-fat diets, the researchers pooled results from randomized controlled trials in adults: participants were assigned to the different diets by chance. That differs from observing people who chose their own diets. Random assignment helps define the comparison, but it does not ensure that everyone eats exactly what was prescribed. The distance between a planned diet and an actual diet becomes important when considering adherence.[1][17]

Even the definition of high protein varies. In another systematic review and meta-analysis involving adults with excess body weight, increased-protein interventions supplied 18–59% of total energy from protein. Across 37 studies assessing body weight, participants consuming more protein lost an average of 1.6 kg more than controls. The 95% confidence interval was 1.2–2.0 kg. Here, that interval indicates uncertainty around the additional weight loss estimated across the increased-protein comparisons.[5]

That average brings together a broad range of dietary conditions. It is not a forecast for a particular dinner or a particular person. To interpret it, ask what the comparison group ate, whether energy intake was restricted, and how far apart the protein intakes were. A diet name is a useful heading, but it does not give the full instructions for assembling a meal.[1][5]

High-protein versus standard-protein low-fat diets at equal energy

The meta-analysis comparing high-protein, low-fat diets with standard-protein, low-fat diets included 24 trials and 1,063 participants. The average diet intervention lasted 12.1 weeks. Both groups restricted energy intake, and researchers prescribed equal energy and matched fat intake while changing the protein-to-carbohydrate balance.[1]

The higher-protein groups lost an average of 0.79 kg more body weight and 0.87 kg more fat mass than the comparison groups. The 95% confidence intervals for the differences were −1.50 to −0.08 kg for body weight and −1.26 to −0.48 kg for fat mass. These are differences between diets, not the total amount lost by the higher-protein participants. There are two quantities to distinguish: the weight each group lost during the intervention, and the additional loss associated with the higher-protein allocation. The reported 0.79 kg belongs to the second comparison. It describes how far apart the groups were on average, rather than where either group started or finished on the scale.[1]

The distinction changes what a reader can reasonably expect. Saying that a diet produced an additional average loss of 0.79 kg does not tell us that its participants lost only that amount. Nor does it establish that someone starting the diet will lose that much more. The result describes the pooled comparison under the conditions of the included trials.[1]

Body weight includes more than fat. In this analysis, the reduction in fat-free mass was smaller in the higher-protein groups by an average of 0.43 kg, with a 95% confidence interval of 0.09–0.78 kg. Fat-free mass means the body’s mass excluding fat; it does not mean muscle alone. Describing this result as an amount of muscle gained would change the finding.[1]

A modest difference on the scale can therefore coexist with a difference in what was lost. The higher-protein diets also produced a greater average reduction in triglycerides, a blood-fat measure, of 0.23 mmol/L. The 95% confidence interval for that difference was −0.33 to −0.12 mmol/L. The decrease in resting energy expenditure was also smaller: the between-group difference was 595.5 kJ/day, with a 95% confidence interval of 67.0–1124.1 kJ/day.[1]

Other measures did not show the same advantage. Changes in fasting glucose, fasting insulin, blood pressure, and total, LDL and HDL cholesterol were similar between the diets. These results belong alongside the favorable findings. An effect on weight or triglycerides cannot be extended to every health measure assessed in the same analysis.[1]

For an everyday packed lunch, the corresponding question is not simply whether the protein portion increased. It is what changed elsewhere in the box: the rice portion, the fat used in cooking, and the overall meal. This analysis tested a change in nutrient allocation within matched dietary conditions. It did not test whether a particular food deserves to win every lunch comparison.[1]

The small average differences should not be dismissed as no effect. They also do not fit an expectation that protein alone creates a large weight-loss advantage. The evidence supports a modest additional benefit for some outcomes when energy restriction is already part of the plan.[1]

The lowest-carbohydrate option does not lead every ranking

A systematic review and meta-analysis examining differences within high-protein diets included 83 articles. Compared with normal-protein diets, higher-protein diets produced an average fat-mass difference of −0.64 kg and a lean-body-mass difference of 0.34 kg. Their 95% confidence intervals were −0.98 to −0.29 kg and 0.11–0.57 kg, respectively.[3]

The researchers used both direct comparisons and a network meta-analysis. A network analysis can compare several diets using evidence from direct comparisons and from shared comparison diets. In the authors’ analysis, a high-protein, moderate-carbohydrate, high-fat pattern ranked highest for outcomes including body weight and fat mass.[3]

For triglyceride reduction and increases in HDL cholesterol, often called “good” cholesterol, the highest-ranked pattern was high protein, low carbohydrate and high fat. The leading diet changed with the outcome. A ranking for triglycerides should not be substituted for a ranking for body weight or body fat simply because the diets share a name.[3]

A ranking also leaves a practical question unanswered: how large was the difference between the leading option and the next one? The reported classifications and rankings do not provide enough information to express that gap in kilograms. They therefore cannot, by themselves, establish how much fat a reader should add to dinner or what individual nutrient proportions to aim for.[3]

The low-fat analysis and the analysis in which a high-fat pattern ranked highest had different comparison conditions. In the low-fat comparison, fat was held at a matched level while protein and carbohydrate proportions changed. In the other analysis, differences in carbohydrate and fat were themselves part of the classification. A favorable result for a high-fat category therefore answers a different question from the modest advantage found when fat was matched. The former matched fat and examined the balance between protein and carbohydrate. The latter classified the remaining macronutrient composition within high-protein diets. Reading them together means keeping track of which question each analysis answered.[1][3]

The feature on ketogenic diets takes a closer look at how their weight-loss results change over time. For this comparison, the key distinction is that high protein and low carbohydrate are not synonyms. When examining a diet, read the instructions for carbohydrate and fat as carefully as the instructions for protein. Otherwise, quite different interventions can appear to be the same choice.[1][3]

Units need attention too. The body-weight result in the analysis of 83 articles was expressed as a standardized mean difference: −0.25, with a 95% confidence interval of −0.40 to −0.11. This is an effect measure expressed on a standardized scale. Unlike the fat-mass result, it cannot be read by attaching “kg” to the number.[3]

The pooled advantage over normal-protein diets and the ranking of different high-protein patterns are separate pieces of information. An advantage in the first comparison does not establish that the highest-ranked pattern in the second comparison is the best fit for a reader’s life. A diet bought at lunchtime and one prepared at home may be difficult to reproduce in the same way. Rankings are a starting point for examining those conditions, rather than ready-made menus.[3]

Numbers: comparing high-protein diets

24 trials, 1,063 participants: the low-fat comparison matched prescribed energy and fat intake.[1]

−0.79 kg in body weight: the average between-group difference, not total weight loss.[1]

−0.64 kg in fat mass: the pooled difference in the analysis examining varied macronutrient patterns.[3]

0.43 kg less fat-free-mass loss: a preservation result that should not be relabeled as muscle gain.[1]

1.6 kg greater weight loss: the estimate across 37 studies in adults with excess body weight, under varied dietary conditions.[5]