368 ‒ The protein debate: optimal intake, limitations of the RDA, & whether high-protein is harmful

KEYPOINTS

• 01

  Protein is the latest macronutrient to be demonized, in a predictable cycle of food villains.

  Obesity researcher and methodologist David Allison (now at Baylor's Children's Nutrition Research Center) says food carries culture, identity, religion and economics, so people keep hunting for 'angels and demons' among fat, carbs, seed oils, phytoestrogens and now protein. Enthusiasm for protein is driving sales, which provokes people who dislike others 'taking the easy way out'. Both he and Peter disclose ties to David Protein; Allison also lists beef and pork industry funding.

• 02

  The 0.8 g/kg RDA was a nitrogen-balance survival threshold, not an optimum.

  It came from nitrogen balance studies in lean, sedentary young men weighing about 65–70 kg. Nobody claimed it was ideal or an upper limit. A 1928 Polish study kept two young adults in nitrogen balance for 6 months on almost nothing but potatoes, without diabetes or weight gain, but that says nothing about older adults, pregnancy, injury recovery or training. Allison's conclusion: almost nobody should aim for just the RDA.

• 03

  Allison's target for thriving: about 2 g/kg/day, spread across meals.

  He sees three zones: too little; a 'known superior' zone up to roughly double the RDA, with no evidence of harm in any group except rare cases like PKU or a specific protein allergy, even in kidney disease; and above about 2 g/kg, more uncertainty and diminishing returns but no demonstrated downturn. Peter's clinic advises 1.6–2 g/kg (about 1 g per pound), and settles for about 1.2 for lifelong vegetarians who struggle to get more. Don Layman would add roughly 30 g per meal.

• 04

  'Trustworthy' depends on data, methods and logic, not on who funded the work.

  Allison separates being trusted (others' judgment) from being trustworthy (process). Critics without good arguments turn to ad hominem attacks and innuendo. He argues scientists should aim to build trust in the scientific process, even at the cost of losing particular arguments.

• 05

  Nutrition trials are chronically underpowered because food isn't patentable.

  Drug trials enroll tens of thousands because patents and FDA requirements justify billions in spending; nutrition RCTs are tiny by comparison. Allison guesses total US spending on research into the effects of food, across all commodity groups and companies, is under $1 billion a year. That's why most nutrition scientists piece together industry and government funding. He'd shift money from expensive observational epidemiology, which yields little new information, to randomized trials.

• 06

  Studies showing a plateau in muscle protein synthesis are misread when applied to trained people.

  Untrained people max out muscle protein synthesis at lower intakes, just as a sedentary person gains hugely from three 30-minute workouts a week that would do almost nothing for Peter. Allison's analogy: using such data to cap protein is like telling a student 10 minutes of algebra a day is enough to master it. The ICU parenteral-nutrition trials showed no mortality benefit from more protein, but no harm either, and don't generalize to people shopping for dinner.

• 07

  No compelling human evidence shows higher protein causes harm, so the burden of proof should shift.

  Observational studies point both ways and protein intake is heavily confounded by social class. Allison also distrusts mouse data: benefits of protein and calorie restriction show up mostly at 22 °C, which is chronic cold stress for a mouse, rather than at thermoneutral 27–30 °C. His black-swan argument: after searching thousands of ponds with drones and binoculars without finding a black swan, it's reasonable to stop worrying about them for practical purposes.

• 08

  'Ultra-processed' is a social construct that's too broad to describe what a food does in the body.

  The NOVA classification lumps meal-replacement shakes, Big Gulps and IV nutrition together. Allison quotes food scientist Joe Schwarcz: a substance's effect 'depends on its molecular structure, not its ancestry'. Nearly everything is processed (pasteurized milk, cheese, wine), and 'natural' foxglove and hemlock are poisons. Avoiding ultra-processed food might work as a rough rule of thumb, like shopping the perimeter of the store, but it says nothing about cause. He prefers David Kessler's 'ultra-formulated'.