Nutrition research and chronic disease

• Nutrition science has not stopped obesity or diabetes because the central drivers remain unresolved.
• The main task for NIH-funded nutrition work is to identify causes of the obesity, diabetes, and related chronic-disease epidemics.
• Nutrition sits under nearly every chronic and many acute diseases, so weak nutrition work weakens health policy across many areas.
• Nutrition trials need more funding because food studies are harder than drug trials: diet adherence, food environment, and behavior all matter.
• A single phase-three drug program can cost about a billion dollars, while nutrition grants are spread thinly across many smaller projects.

How bias enters the field

• Longstanding belief systems favor eating less fat, avoiding saturated fat, and eating mostly plants, whole grains, legumes, fruits, and vegetables.
• A study that fits those beliefs can win prestige even when its design is weak.
• A study that conflicts with those beliefs can struggle for attention even when it asks an important question.
• Confirmation bias exists on all sides, including low-carb advocates, but glaring statistical flaws still pass when the result fits the dominant view.

Short feeding studies

• Inpatient feeding trials can look impressive because food is controlled, subjects are watched, and metabolism is measured with advanced tools.
• The problem is the timescale: chronic diseases develop over years, while two-week trials mainly capture metabolic transition.
• Low-carb adaptation takes several weeks because the brain moves from glucose dependence toward ketone use.
• The first two weeks of a low-carb diet can include fatigue, hunger, and "keto flu," so early data can confuse adaptation with chronic-disease effects.

Kevin Hall low-fat versus ketogenic trial

• The 2021 Nature Medicine trial compared a plant-based low-fat diet with an animal-based ketogenic diet in a four-week crossover feeding design.
• Subjects ate one diet for two weeks, then immediately crossed to the other diet without a washout period.
• The original paper found lower calorie intake on the low-fat diet and was read as a strike against the carbohydrate-insulin model.
• Later diet-order analysis found a massive carryover effect, roughly 2,000 calories per day.
• That carryover undermines the original low-fat-versus-keto calorie conclusion because the second period was contaminated by the first period.
• The responsible fix is correction, reanalysis, or retraction, because the paper remains in the literature as low-fat evidence.

Ultra-processed food

• Ultra-processed food is a broad and recent concept built around processing, additives, and whether a food can be made in a home kitchen.
• The category can group unlike foods together: Coca-Cola versus homemade lemonade, and multi-ingredient ice cream versus simple ice cream.
• Processing matters more for carbohydrates than for fats and proteins in this account: wheat berries to white bread and fruit to juice change insulin dynamics more than olives to olive oil or steak to hamburger.
• Additives are not one thing: some are innocuous or helpful, while emulsifiers can disrupt the gut lining.
• Epidemiology links higher ultra-processed-food intake with worse health, but the heavier consumers also have lower income, less exercise, more smoking, and other confounders.
• The trial base is too thin if public policy rests on one two-week trial and one one-week trial.

Better trial designs

• A useful trial would compare low-carb, low-ultra-processed, and low-fat diets in parallel groups for at least one year, ideally two.
• Each group needs enough support: dietitians, in-home counseling, and possibly provided food.
• The low-carb arm should be meaningfully low-carb, around 25 percent carbohydrate or less, not gradually liberalized into a higher-carb diet.
• The low-ultra-processed arm should focus on avoiding ultra-processed foods while still allowing carbs such as potatoes, grains, and homemade desserts.
• The low-fat arm can serve as a conventional comparator.
• A trial like this might cost around $20 million, still far below a phase-three drug trial.

DIETFITS and low-carb evidence

• DIETFITS found little difference between healthy low-fat and healthy low-carb diets.
• A later reanalysis found results favoring low-carb and matching the carbohydrate-insulin model, with high insulin secretors more sensitive to carbohydrate load.
• DIETFITS reduced sugar and glycemic load in both groups, which narrowed the dietary contrast.
• DIETFITS also liberalized carbohydrate intake over time, so the low-carb arm no longer stayed very low-carb.
• Self-selected diet studies cannot cleanly answer causal questions because people choosing different diets differ from the start.

Low-carb conditions and institutional inertia

• Carbohydrate restriction has a long history in type 2 diabetes, obesity, and pediatric epilepsy.
• Public institutions still have not run large NIH-scale, multicenter low-carb trials comparable to major low-fat trials.
• Look AHEAD spent major resources on a low-fat lifestyle intervention in type 2 diabetes and stopped early for futility on heart-disease outcomes.
• Low-carb adherence can become a self-fulfilling failure when researchers assume people cannot follow it and then provide little support.
• Patients who feel less hungry and lose weight with modified low-carb diets may gain self-efficacy and continue willingly.

Public health versus precision nutrition

• The likely answer is between one-diet-for-everyone and a fully individualized AI diet for each person.
• Public health still needs identification of the main drivers, such as processed carbohydrates if the carbohydrate-insulin model is correct.
• High insulin secretors, often with central fat gain, may be especially vulnerable to high-carbohydrate diets.
• The Nutrition for Precision Health initiative uses omics, AI, three diets, three periods, and short diet windows to predict personalized diets.
• A two-week diet window may be too brief to determine long-term disease risk or long-term dietary success.
• Ten to seventeen million dollars per long-term trial would be more useful than one $170 million short-term precision-nutrition project.

Bottom line

• Nutrition science needs fewer elegant short trials that cannot answer chronic-disease questions and more long, supported, disease-relevant trials.
• The key question is still causal: what dietary forces created the obesity and diabetes epidemics, and what diets reliably reverse or prevent them?
• Without definitive evidence, public guidance becomes authority-driven messaging, not reliable science.

References

• [02:54] Good Calories, Bad Calories — https://garytaubes.com/works/books/good-calories-bad-calories/
• [02:58] Rethinking Diabetes — https://garytaubes.com/works/books/rethinking-diabetes/
• [03:09] Uncertainty Principles — https://uncertaintyprinciples.substack.com/
• [03:18] NIH Has A Nutrition Problem (Part 2) — https://uncertaintyprinciples.substack.com/p/nih-has-a-nutrition-problem-part
• [03:22] Wash-in and washout effects: mitigating bias in short term dietary and other trials — https://doi.org/10.1136/bmj-2024-082963
• [03:25] Designed to Fail: Why Short-term Diet Trials Cannot Solve the Epidemics of Chronic Disease — https://davidludwigmd.medium.com/designed-to-fail-44d278e591a6
• [07:55] Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake — https://doi.org/10.1016/j.cmet.2019.05.008
• [08:00] Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake — https://doi.org/10.1038/s41591-020-01209-1
• [12:39] Nutrition for Precision Health, Powered by the All of Us Research Program — https://clinicaltrials.gov/study/NCT05701657
• [23:25] Diet order significantly affects energy balance for diets varying in macronutrients but not ultraprocessing in crossover studies without a washout period — https://doi.org/10.1016/j.ajcnut.2024.08.013
• [31:07] Nutrition and health. The issue is not food, nor nutrients, so much as processing — https://doi.org/10.1017/S1368980009005291
• [43:19] Effect of Low-Fat vs Low-Carbohydrate Diet on 12-Month Weight Loss in Overweight Adults and the Association With Genotype Pattern or Insulin Secretion — https://doi.org/10.1001/jama.2018.0245
• [43:44] Evidence for the carbohydrate-insulin model in a reanalysis of the Diet Intervention Examining The Factors Interacting with Treatment Success (DIETFITS) trial — https://doi.org/10.1016/j.ajcnut.2022.12.014
• [50:59] Cardiovascular effects of intensive lifestyle intervention in type 2 diabetes — https://doi.org/10.1056/NEJMoa1212914

:::