Is There a Best Diet? What Head-to-Head Trials Actually Show

Ask the internet which diet is best and you start a war. Keto against vegan, paleo against plant-based, carbs as poison against fat as poison, each camp armed with anecdotes and certainty. Meanwhile, the people caught in the middle cycle from one plan to the next, convinced the next one is "the answer."

In Comparing Diets, I take a calmer route. Instead of crowning a winner, the book gives you a method for comparing diets on their merits, then applies it to every major pattern: Mediterranean, DASH, low-carb and ketogenic, vegetarian and vegan, paleo and Whole30, intermittent fasting, carnivore, and the fads and detoxes that fill the gaps. Here is the core of what the evidence says.

Is there one best diet for everyone?

No. When healthy versions of different diets are compared head-to-head, the average differences are small and the variation between individuals is large.

The clearest example is the DIETFITS trial (Gardner et al., 2018). Researchers assigned 609 adults to either a healthy low-fat or a healthy low-carbohydrate diet for 12 months. Both groups were told to eat whole, minimally processed foods, without counting calories. The result: no significant difference in weight loss, roughly 5.3 kg on low-fat versus 6.0 kg on low-carb, with wide individual variation in both groups.

Two details matter even more. First, neither a person's genotype pattern nor their insulin-secretion status predicted which diet worked better for them, which undercuts many popular "personalization" claims. Second, people succeeded on both diets by doing the same things: cutting added sugar and refined grains, and eating more vegetables and whole foods. Earlier trials such as A TO Z (Gardner et al., 2007) and a broad meta-analysis of named diet programs (Johnston et al., 2014) point the same way. Named diets differ little on average. Adherence and food quality dominate.

What do healthy diets have in common?

Strip away the branding, and the diets with the best evidence overlap enormously. The book calls this overlap the common core:

1. Built on whole, minimally processed foods, with ultra-processed foods kept to a minimum.
2. Plant-rich: plenty of vegetables, fruits, legumes, whole grains, nuts and seeds.
3. Adequate protein, from the sources you prefer.
4. Adequate fibre.
5. Mostly unsaturated fats, from oils, nuts, seeds and fish.
6. Limited added sugars, refined grains and sugar-sweetened beverages.
7. Limited excess sodium and excess alcohol.
8. Attention to portions and energy density.

Michael Pollan's seven-word summary captures it well: "Eat food, not too much, mostly plants" (Pollan, 2008). It is a heuristic, not a prescription, but it describes what the successful DIETFITS participants did on both arms.

The practical consequence is what I call the "healthy version" qualifier. Low-carb can mean grilled fish, eggs, nuts and vegetables, or it can mean processed meats and cheese. Plant-based can mean legumes and whole grains, or chips, soda and vegan junk food. The healthy version of almost any diet converges on the common core. That is why "healthy versus ultra-processed" is a better question than "low-carb versus low-fat."

Do calories still count, or is it all about insulin?

Calories count. Over time, the change in your body's energy stores follows the difference between what you take in and what you expend. Diets that work create a calorie deficit, usually by reducing intake.

The carbohydrate-insulin model proposes the opposite: that carbohydrates raise insulin, insulin drives fat storage, and low-carb diets therefore offer a "metabolic advantage" (Ludwig & Ebbeling, 2018). Carefully controlled tests do not support its strong form. In metabolic-ward studies, where intake is fully controlled, low-fat and low-carb diets produce similar fat loss calorie for calorie. In one short crossover trial of 19 adults spending 6 days on each diet, fat restriction actually produced slightly more body-fat loss (Hall et al., 2015). The book is candid that these trials are short and that the model's proponents dispute how they should be read, so the debate continues at the margins.

But food choice still matters a great deal, just not in the way the insulin story claims. A calorie is a calorie thermodynamically, but not behaviourally. What you eat shapes how much you end up eating. In a metabolic-ward trial of 20 adults (Hall et al., 2019), people ate about 500 kcal more per day on an ultra-processed diet than on a minimally processed one matched for presented calories, macronutrients, sugar, fibre and sodium. They gained about 0.9 kg on the first and lost about 0.9 kg on the second over two weeks each. The trial was small and short, so it shows a direction rather than a long-term effect, but the message is clear: protein, fibre and whole foods help you feel full on fewer calories, while ultra-processed foods make overeating easy.

How can you judge any diet on your own?

A verdict on one diet helps you once. A framework helps with every diet you will meet, including the ones not invented yet. The book scores each pattern on seven criteria:

• Nutritional adequacy: does it meet macro- and micronutrient needs, or does it eliminate food groups and risk deficiencies?
• Energy and weight effect: does it support satiety rather than relying on willpower?
• Health markers: what happens to blood lipids, blood pressure and blood glucose?
• Safety: short- and long-term risks, and interactions with medications or conditions.
• Sustainability and adherence: can you follow it long-term without deprivation and rebound?
• Evidence quality: randomized trials or testimonials, and how long were the studies?
• Practicality, cost and culture: does it fit your budget, skills, time and food traditions?

No diet aces every row. The point is to make the trade-offs visible. Of the seven, adherence is the decisive one: in a classic comparison of four popular diets, people who stuck to their assigned plan lost more weight regardless of which plan it was (Dansinger et al., 2005).

The book also lists red flags of a weak claim: before-and-after testimonials, mechanism-only reasoning ("insulin," "toxins"), demonizing a single food, promises of rapid and effortless results, a product to sell, and appeals to "ancient" or "secret" knowledge.

Which diet should you choose?

The final chapter turns all of this into four steps. Start with the common core, which is non-negotiable. Match the pattern to the person: goal, health status, preferences, culture, budget, cooking skills and values. Give adherence the most weight. Personalize the macronutrient ratio last.

Absent a specific reason to choose otherwise, a Mediterranean-style, whole-food, plant-rich pattern is the book's evidence-based default, and it adapts to most goals and cultures. For active people and the coaches who work with them, the same logic applies: protein and fibre for satiety, small changes over dramatic overhauls, an environment that makes good choices easy, and treating a lapse as a lapse rather than a relapse.

One boundary stays firm throughout. Diets used to treat disease, dietary changes made while taking glucose-lowering or blood-pressure medication, therapeutic elimination diets, pregnancy and childhood nutrition, and any form of disordered eating call for individual care from a physician or registered dietitian.

Read the full comparison

If you want every pattern scored on the same seven criteria, from DASH and keto to intermittent fasting, carnivore and detox claims, the full analysis is in Comparing Diets, part of The Performance & Human Physiology Encyclopedia, the 93-book Netritious collection I wrote for people who train, coach and care for others.

Frequently asked questions

Is low-carb better than low-fat for weight loss?

Not on average. In DIETFITS, healthy low-carb and healthy low-fat diets produced similar weight loss at 12 months (Gardner et al., 2018). The better choice is the one you can sustain with good food quality.

Is intermittent fasting more effective than regular calorie restriction?

The book treats intermittent fasting and time-restricted eating as roughly equivalent to calorie restriction: a convenient structure for some people, not a metabolic shortcut.

Do detox diets remove toxins?

No. Your liver and kidneys do that work. The book classifies detoxes and cleanses as marketing rather than a proven health practice.

Where can I find the book Comparing Diets?

Comparing Diets is available on Amazon in Kindle format, in English.