20–25% Energy from Protein Is the Best Weight Maintenance Zone: NoHoW Data

In an analysis of 1,518 adults who had already lost at least 5% of their weight, the researchers showed that the lower the proportion of energy from protein (and the diet was “diluted” with fat and/or carbohydrates), the higher the overall appetite and energy consumption - and the more noticeable the return of weight and the increase in central obesity over 12 months. The main “culprit” is discretionary products (sweets, fast food, snacks, alcohol, etc.): they reduce the percentage of protein in the diet, pushing people to overeat. Conversely, maintaining the percentage of protein within the “normal corridor” helps maintain the result. The work was published in the journal Obesity.

Background

• Weight maintenance is a weak link in most programs. Even after clinically significant weight loss (≥5%), it is difficult to maintain the result for a year, which was the focus of the European randomized NoHoW project (UK/Denmark/Portugal), which tested digital weight management tools and collected detailed food diaries and objective activity metrics. This data set formed the basis of the current analysis.
• Which "balance of BJU" is better for weight maintenance is an open question. The "fat vs. carbohydrate" debate has been going on for decades and yields contradictory results; much indicates that it is not only caloric content that is important, but also the ratio of macronutrients and their impact on appetite and energy metabolism.
• "Protein leverage": the key idea. Back in 2005, Stephen Simpson and David Raubenheimer proposed the hypothesis that humans tightly regulate protein intake. If the proportion of protein in the diet is diluted with fat/carbohydrates, the body tends to "get" protein, which leads to overeating and increased energy - with a relatively stable absolute protein intake. The model is supported theoretically and by data from various species.
• Observational evidence supports the “lever” approach. Dietary analyses in the US (NHANES 2009–2010) and Australia have shown that the higher the proportion of ultra-processed/discretionary foods, the lower the percentage of energy from protein and the higher the total energy intake, while absolute protein remains roughly constant. The “nutrient geometry” approach clearly demonstrates these relationships.
• Randomized Evidence: Slightly More Protein, Less Rogaine. In the large European DIOGenes trial, the combination of moderately increased protein and a low glycemic index diet improved weight maintenance after weight loss; the effect was replicated in follow-up and follow-up analyses.
• What's different about the new work in Obesity (2025) is that for the first time, using a large WLM cohort from NoHoW (n=1,518), the authors correlated protein percentage with 12-month changes in body weight and abdominal indices using mixture models and feeding geometry. They found that low protein % (often due to excess discretionary foods) was associated with higher energy and greater weight return/waist-to-height gain, while maintaining protein % was associated with better retention.
• Practical context. The results fit into a broader picture: when maintaining weight, it is not "high protein at any cost" that is important, but rather preventing "protein dilution" through sweets, snacks, fast food and alcohol - it is through a drop in protein percentage that these products "turn on" the protein lever and push towards extra calories.

What was studied

• Data from a 12-month follow-up of participants in the European NoHoW project (UK, Denmark, Portugal), where all participants had already successfully lost weight (≥ 5%) during the previous year and were trying to maintain their weight. Diet was recorded with 4-day 24-hour reminders; then “nutrient geometry” was applied - models that consider the ratio of the three macronutrients as a total mixture (the percentages of energy from protein, fat and carbohydrates are summed up to 100%).

Key Results

• On average, the participants' diets contained ~21% of energy from protein, 34% from fat, and 43% from carbohydrates. But the lower the protein content, the higher the food mass and daily energy; statistically, this was expressed by the energy correlation coefficient β = −0.33: as the protein content increased, the energy decreased, and vice versa.
• On the response surfaces, areas of low protein coincided with worse retention trajectories: greater weight regain, increased waist-to-height ratio, and increased hip-to-height ratio over 12 months. No clear association was found for fat mass index (FMI).
• Discretionary foods (energy-intensive, rich in saturated fats/sugars/salt/alcohol, poor in fiber) had a median of ~4% protein and were located in the "low-protein" zones of the triangle. The more of them in the diet, the lower the proportion of protein and the higher the total energy. If, however, the model is statistically "clamped" by the percentage of protein, the "discretionary - energy" relationship itself disappeared - that is, it is the dilution of protein that explains overeating.

Why it matters: The "protein lever"

The idea of protein leverage has been discussed for a long time: people strive to satisfy their protein needs, and if the diet is poor in protein (even if it has a lot of calories), the body “squeezes out” the appetite - we eat more to get protein, while overeating fat and carbohydrates. New work transfers this principle to the context of long-term weight maintenance: not the proportion of fat vs. carbohydrates, but the percentage of protein became the strongest predictor of energy and regeneration.

What are considered "discretionary" products?

Sweet pastries and desserts, candies, sweet drinks, alcohol, chips and other snacks, fast food, some processed meat products. Their common feature is little protein, a lot of easily accessible energy and low satiety per calorie. It is by reducing the percentage of protein that they "turn on" the protein lever and push towards extra calories.

Restrictions

• This is a secondary analysis of the WLM cohort with digital intervention; causal inferences are cautious.
• Self-reporting of dietary intake is always vulnerable to underreporting, but the authors conducted sensitive analyses (including under-eating/under-reporting scenarios) and the main findings held up.
• The sample is European; generalization to other countries requires verification.

What to do for a reader who is holding weight

• Keep the protein percentage in the "corridor" (guideline - ≈ 18-25% of energy; the exact range is selected with a nutritionist). This is about the percentage, not about "eat as much protein as possible."
• Reduce discretionary foods: they tend to "dilute" protein and stimulate appetite. Replace them with foods from the "five groups" (fish/eggs/tofu/lean meat and dairy, whole grains, vegetables, fruits, nuts/seeds) - they have a higher proportion of protein and are more filling per calorie.
• Don't get hung up on "fat vs. carbohydrates": the structure of the diet as a whole and the percentage of protein are more important than this dualism. And, of course, physical activity and self-monitoring (weighing, steps) remain the basis for retention.

Source: Zhang H, Vasileiou A, Searle D, Larsen SC, Senior AM, Magkos F, et al. Dietary Macronutrient Composition and Protein Concentration for Weight Loss Maintenance. Obesity (published online 7 August 2025). https://doi.org/10.1002/oby.24370