What's the Healthiest Diet? Cutting Through the Conflicting Advice

What we checked — and why the answer is less controversial than it seems

The confusion around diet advice is real, but it's largely manufactured. We looked past the headlines — past the industry-funded studies, the single-nutrient scares, and the branded diet books — and went directly to the large, long-running bodies of evidence: controlled trials, decades-long cohort studies, and systematic meta-analyses involving hundreds of thousands of people. The signal that keeps emerging from that data is consistent, and it points in one direction.

We also looked specifically at where official dietary guidelines have lagged behind or diverged from the evidence — particularly the decades-long under-emphasis on ultra-processed food as a category, and the over-emphasis on individual nutrients like dietary fat and cholesterol. Where the evidence contradicts the guidelines, we follow the evidence.

• ✓
  Long-term randomised controlled trials reviewed The PREDIMED trial (7,447 participants, ~5 years) found the Mediterranean diet reduced major cardiovascular events by roughly 30% compared to a low-fat control diet — one of the largest diet RCTs ever conducted.
• ✓
  Large cohort studies cross-referenced The Nurses' Health Study, Health Professionals Follow-Up Study, and EPIC cohort — collectively tracking hundreds of thousands of people for decades — consistently show whole-food, plant-forward diets associated with lower all-cause mortality.
• ✓
  Ultra-processed food evidence assessed Research published in The BMJ and The Lancet shows ultra-processed food consumption is independently associated with higher rates of obesity, type 2 diabetes, cardiovascular disease, and cancer — regardless of the overall macronutrient breakdown of a diet.
• ✓
  Low-carb and keto claims examined Short-term trials show low-carb diets can accelerate initial weight loss, but no large long-term RCT demonstrates superior outcomes for mortality or disease prevention compared to a Mediterranean-style whole-food diet.

The evidence supports one base pattern — but there's real flexibility in how you get there

The research doesn't mandate a single rigid meal plan. What it does establish is a base pattern. How you implement that pattern can reasonably vary based on your health conditions, food culture, budget, and personal preferences.

Diet approaches that are popular but poorly supported by the long-term evidence

Most of the conflicting advice you've encountered comes from ideas that are plausible-sounding, aggressively marketed, or rooted in short-term data — but that don't hold up when you look at what actually happens to people over years and decades.

• ✗
  Chasing macronutrient ratios above all else — Decades of research framing dietary fat, then carbohydrates, then protein as the single villain responsible for poor health has repeatedly failed to produce lasting public health improvements; what matters far more than macro ratios is the overall quality and degree of processing of the foods you eat.
• ✗
  Ultra-processed "diet" or "health" food products — Low-fat diet yogurts, protein bars, meal-replacement shakes, and keto snack packs are still ultra-processed foods; they carry the same metabolic and inflammatory signals as other processed products regardless of the macronutrient profile printed on the front of the packaging.
• ✗
  Detoxes, cleanses, and elimination protocols without a clinical basis — There is no credible mechanism by which a juice cleanse or multi-week elimination protocol "detoxifies" the body; your liver and kidneys do this continuously, and these protocols typically cause short-term weight loss through calorie restriction alone, with no lasting benefit.
• ✗
  Treating a single superfood as transformative — No individual food — not blueberries, not turmeric, not chia seeds — has meaningful impact on long-term health in isolation; the benefit attributed to these foods in observational studies almost always reflects the broader dietary pattern of the people eating them.
• ✗
  Relying on short-term study results reported in the press — A study showing a 12-week dietary intervention improved a biomarker in 80 participants tells you almost nothing about what you should eat for life; the press reports these studies as if they overturn decades of evidence, and diet-book authors build entire brands on them.