#368 ‒ The protein debate: optimal intake, limitations of the RDA, whether high-protein intake is harmful, and how to think about processed foods | David Allison, Ph.D.
David Allison, a renowned obesity researcher, discusses the contentious topic of protein, dissecting the RDA's origins and limitations, the challenges of nutrition science, and the debate around processed foods. He also explores public health failures in tackling obesity and the potential of GLP-1 agonists.
Deep Dive Analysis
14 Topic Outline
The Historical Cycle of Demonizing Macronutrients
Origins and Limitations of the Protein RDA
Trustworthiness in Science: Data, Methods, and Logic
Challenges of Nutrition Science and Crossover Designs
Critiquing the Protein RDA for Optimal Health
The Absence of Evidence for Harm from High Protein Intake
Underfunding of Nutrition Trials vs. Pharmaceutical Trials
Limitations and Biases of Nutrition Epidemiology
Practical Protein Intake Targets for Thriving
Defining Processed and Ultra-Processed Foods
The Molecular Structure Principle of Food Effects
Why Conventional Public Health Interventions for Obesity Fail
Radical Ideas for Addressing Metabolic Health
The Potential Role of GLP-1 Agonists in Public Health
7 Key Concepts
Data, Methods, and Logic
These are the three fundamental components that determine the trustworthiness of scientific findings. The data collected, the methods used to collect them, and the logical connection between the data and conclusions are paramount, while other factors like funding sources are tangential.
Crossover Designs
A study design where subjects receive a sequence of different treatments (e.g., diet A then diet B). While statistically powerful, they have an Achilles heel of 'carryover effects,' where the first treatment might influence the effects of the second, even with a washout period.
Carryover Effects
In crossover study designs, this refers to the lingering influence of a prior treatment on the effects of a subsequent treatment. This can confound results, making it difficult to isolate the true effect of each intervention.
Statistical Power
The probability of correctly rejecting a false null hypothesis (i.e., detecting a true effect if one exists). Crossover designs typically offer greater statistical power, allowing researchers to use fewer subjects to achieve similar precision compared to parallel group designs.
NOVA Classification
A widely used, though controversial, system for classifying foods based on the degree and purpose of processing. It categorizes foods into groups ranging from unprocessed to ultra-processed, but lacks a single, universally accepted definition.
Molecular Structure Principle
This principle states that the effect of substances in the body depends on their molecular structure, not their ancestry or origin. Whether a molecule is naturally occurring or synthetically produced, its biological effect is determined by its chemical composition and structure.
Ultra-Formulated Foods
A term proposed by David Kessler to describe foods engineered for hyper-palatability, often by combining fat, sugar, and salt in specific ways. This term focuses on the composition and design of the food rather than just the processing steps, which is considered a more meaningful classification than 'ultra-processed.'
10 Questions Answered
Nutrition science often cycles through demonizing different macronutrients (like fat, then carbs, now protein) and specific foods (like seed oils or phytoestrogens), driven by economic interests, cultural identity, and the search for villains and heroes in food.
The RDA originated from early 20th-century nitrogen balance studies on lean, inactive young men, showing it was enough to maintain nitrogen balance and compatible with survival. However, it was never claimed to be optimal or an upper limit, and it doesn't account for different goals like muscle building, recovery, or aging.
No, for most people aiming to thrive, optimize health, or achieve goals like avoiding sarcopenia, improving physical performance, or building muscle, the RDA is likely insufficient. It represents a minimum for survival, not an optimal intake for robust health.
Challenges include methodological difficulties in collecting accurate data (e.g., measuring food intake in free-living people), the high cost of rigorous trials, issues with adherence and duration, the emotional and social aspects of food, and the difficulty of blinding dietary interventions.
There is no compelling evidence from human intervention studies showing deleterious effects of higher protein intake (up to 2 g/kg/day or more) on clinically meaningful endpoints like cancer or heart disease, even in populations like those with chronic kidney disease, with rare exceptions for specific amino acid disorders.
Pharmaceuticals are patentable, allowing companies to recoup massive R&D investments (hundreds of millions to billions of dollars) required for FDA approval. Food products are generally not patentable, leading to lower economic incentives for the food industry to fund large, rigorous, and expensive nutrition studies.
Key limitations include confounding (especially by culture, socioeconomic status, and education), non-random measurement error in self-reported food intake, and selection biases. Additionally, there can be issues with the sincerity of reporting by investigators, leading to distortion or manipulation of information.
There isn't a single, universally accepted definition, but systems like NOVA classify foods based on the number and type of processing steps involved. 'Processed' generally refers to any alteration (e.g., cutting, drying, cooking), while 'ultra-processed' implies more extensive industrial formulation, often with many ingredients and additives, commonly found in packaged goods.
The concept of 'ultra-processed' is often used as a heuristic (a simple rule of thumb) to guide dietary choices, such as avoiding foods from the center aisles of a grocery store. However, from a scientific perspective, the category itself is considered too broad and not intrinsically meaningful, as the effects of food depend on its molecular structure and composition, not its processing history or 'ancestry.'
Public health interventions for obesity face intrinsic difficulties because eating is essential for survival, unlike smoking cessation. People value freedom and variety in food choices, making broad restrictions challenging. Many 'obvious' interventions (school-based programs, walking trails, menu labeling) have not shown large, demonstrable, meaningful effects, suggesting a need for more radical and creative approaches.
14 Actionable Insights
1. Increase Protein Intake
Aim for a protein intake in the neighborhood of 1.6 to 2 grams per kilogram of body weight per day, as there is strong evidence for benefits in body weight, appetite control, bone strength, and muscle, with no known evidence of harm for most people.
2. Optimize Protein, Don’t Just Pass
View the protein RDA as a ‘pass’ grade for survival, but if your goal is to ’thrive’ by being stronger, healthier, and living longer, aim for a higher intake (around 2 grams per kilogram per day) to optimize outcomes.
3. Distribute Protein Intake
Aim to eat protein at least three or four times a day, with approximately 30 grams at each sitting, rather than consuming a large amount in one go, to optimize muscle protein synthesis.
4. Rethink Protein RDA
Understand that the recommended daily allowance (RDA) of 0.8 grams of protein per kilogram of body weight is based on studies of lean, inactive young men and is likely only sufficient for survival, not for optimizing health, performance, or preventing sarcopenia.
5. Align Nutrition with Goals
Recognize that general nutritional recommendations, like the protein RDA, are often based on specific populations and goals (e.g., survival), so adjust your intake based on your own goals such as slowing aging, building strength, or recovering from injury.
6. Challenge Protein Harm Claims
Be skeptical of claims that higher protein intake is harmful unless presented with controlled intervention studies in humans showing deleterious effects on clinically or intrinsically meaningful endpoints, as such evidence is currently lacking.
7. Prioritize Food’s Molecular Structure
Understand that the effect of substances in the body depends on their molecular structure, not their ancestry (e.g., whether it’s ’natural,’ ’locally grown,’ or ‘processed’), so focus on the actual chemical composition of food.
8. Question ‘Ultra-Processed’ Labels
Recognize that ‘ultra-processed’ is a controversial and broad category that lacks a single accepted definition, and includes vastly different products, making it an unhelpful term for understanding the intrinsic effects of food.
9. Shop Grocery Store Periphery
As a heuristic, try to buy most of your food from the periphery of the grocery store (dairy, produce, meats, fish) and less from the center aisles, as this often correlates with consuming fewer ultra-processed and energy-dense foods.
10. Assess Scientific Trustworthiness
When evaluating scientific claims, focus on the data presented, the methods used to collect that data, and the logic connecting the data to the conclusions, rather than tangential factors like funding sources or ad hominem attacks.
11. Acknowledge Science Limits
Understand that nutrition science faces significant challenges in data collection, blinding, adherence, and duration, meaning that many findings offer ‘rough inferences’ and ‘supported answers’ rather than absolute, unequivocal proof.
12. Understand Study Design Flaws
Recognize that study designs like crossover trials, while statistically powerful, can have limitations such as ‘carryover effects’ where previous treatments might influence subsequent results, which can weaken the certainty of conclusions.
13. Invest in Broad Social Research
Advocate for research funding into the long-term effects of general education, financial security, and quality of upbringing on obesity levels, rather than focusing solely on nutrition education or trivial variants of past public health interventions.
14. Evaluate GLP-1 as Default
As GLP-1 agonist-related drugs continue to demonstrate profound benefits, consider the future possibility that they might become a default public health intervention, similar to childhood vaccines or fluoridated water, for a broad population.
5 Key Quotes
In science, three things matter: the data, the methods used to collect the data, which give them their probative value, which shows what they mean, and the logic connecting the data to conclusions. And everything else is tangential.
David Allison
If I look at a thousand swans and I cannot find a black swan, does that mean that no black swans exist? Of course not. But if I looked at 10,000 black swans, if I sent a helicopter in the air and we scoured and we looked with binoculars, if I sent teams of undergraduate students to walk around ponds and measure, look for every black swan, if I put drones with cameras out and I have failed time after time to find a black swan at one point and say, maybe for practical purposes, I can say black swan's probably not something we need to worry about.
David Allison
The whole purpose of eating is to get chemicals into the body, to replace the chemicals the body loses through the process of living. All food is chemicals. We are chemicals.
David Allison
The effect of substances in the body depends on their molecular structure, not their ancestry.
Joe Schwartz (quoted by David Allison)
This is a heart attack on a plate.
Michael Jacobson (quoted by Peter Attia)
1 Protocols
General Protein Intake for Thriving
David Allison- Aim for approximately 2 grams of protein per kilogram of body weight per day.
- Distribute protein intake throughout the day, ideally with multiple servings (e.g., 3-4 times a day).
- Ensure each protein serving is substantial (e.g., around 30 grams of protein per sitting) to maximize benefits.