Nutrition appears to be a field of constant reversal, which is partly true of the frontier and not true of the foundations, where agreement has been stable for decades.
Where the consensus is
Eating more vegetables, fruit, pulses, nuts and whole grains is associated with better outcomes across essentially every framework.
Which appears in guidance from every major health body, in every dietary pattern with evidence behind it, and across observational and trial evidence.
High intakes of processed meat and of foods high in added sugar are associated with worse outcomes, similarly consistently.
Excess alcohol is harmful, with recommended limits having fallen as evidence for protective effects weakened under better analysis.
And adequate protein, adequate micronutrients and adequate total energy matter, with deficiency causing identifiable problems.
Where the arguments are
The precise role of saturated fat, where evidence has been reassessed and the picture is more nuanced than earlier guidance.
Optimal macronutrient ratios, where trials comparing low-fat and low-carbohydrate approaches generally find similar outcomes when adherence is matched.
Specific foods, where individual studies produce headlines that reverse with the next study.
Which is where the noise is, and it concerns a small fraction of what actually determines dietary quality.
The adherence finding
The most useful result in the whole field.
Trials comparing dietary approaches consistently find that adherence predicts outcomes more than the specific approach does.
Which means the best diet is largely the one a person will actually maintain, and that depends on preference, culture, cost and circumstance rather than on optimisation.
It also explains why every approach has enthusiastic advocates — each worked for the people who stuck to it.
Dietary patterns
Research has moved from individual nutrients toward whole patterns, which is more informative since people eat food rather than nutrients.
The patterns with the best evidence share features — plant-heavy, minimally processed, moderate in animal products — while differing substantially in specifics.
Which supports the view that several quite different diets are compatible with good outcomes, and that the shared features matter more than the differences.
Processing
Currently the most active area, with consistent associations between highly processed foods and poor outcomes.
The mechanism is debated — whether it is the processing itself, the resulting energy density and palatability, or the displacement of other foods.
A controlled feeding trial found higher energy intake on a processed diet than on a matched unprocessed one, which is suggestive and comes from a single small study.
What to ignore
Single studies reported as reversals, since the evidence base moves slowly and single studies rarely change it.
Detox claims, since the organs responsible for this function do it continuously and are not assisted by any product.
Superfood framing, since no single food changes outcomes meaningfully against a background diet.
And elimination of entire food groups without a diagnosed reason, which reduces dietary variety for no established benefit.
The practical version
Eat more plants. Cook more of what you eat. Reduce highly processed foods and added sugar. Drink less alcohol. Eat enough protein.
Which is unexciting, is what the evidence supports, and would be the answer regardless of which specific argument is currently being had.
Anyone with a medical condition, or considering a substantial dietary change, should speak to a doctor or a registered dietitian rather than acting on general commentary.
Cooking at home
Associated with better dietary quality across many studies, largely because it displaces prepared and eaten-out food.
Which is one of the more actionable findings, and the barriers are time, skill and equipment rather than knowledge.
Cooking skills interventions have shown modest effects, and the time constraint is the harder one.
Cost
Healthier diets generally cost more per calorie and can cost less per weight of food.
Which is a real constraint that dietary advice frequently ignores, and pulses, frozen vegetables, tinned fish and whole grains are among the cheapest sources of what the guidance recommends.
Food insecurity affects dietary quality more than knowledge does for the households experiencing it.
Portion and context
Plate size, packaging size and serving format all affect how much is eaten, in a substantial experimental literature.
Which means environment does considerable work that willpower is credited with.
Changing what is easily available at home is more effective than deciding to eat differently.
Reading claims
Packaging claims are regulated in most markets and the regulated terms are narrower than they sound.
Which means checking the ingredient list and the nutrition panel is more informative than any front-of-pack statement.
Anyone with a specific condition should get individual advice rather than applying population-level guidance.
Sustainability
Dietary patterns with better health outcomes generally have lower environmental impact, since both point toward more plants and less meat.
Which is a convenient alignment rather than a coincidence, and several national guidelines now address both together.
Children
Requirements differ from adults, particularly regarding energy density and specific nutrients during growth.
Which means applying adult guidance to children can be actively harmful, and paediatric advice should come from a clinician.