Diet and Health: An Evidence-Based Reading of Food Order, Processing, and Extreme Diets

Diet and Health: No Universal Miracle Diet, and No Fear-Based Labels

The conclusion in one sentence

A sound approach to healthy eating is to choose a sustainable pattern based on minimally processed, nutritionally diverse foods and adjust it for one’s diseases, medications, and individual circumstances—not to assume that one way of eating suits everyone. Multiple healthy dietary patterns are associated with a lower risk of death and can be adapted to different food traditions and preferences PMID: 36622660

Fruit before or after a meal: first, dispel the “rotting” myth

In his original article, Professor Ching-Shun Lin addressed an online claim by explaining that fruit does not have to be eaten on an empty stomach to be properly absorbed, nor will it rot in the stomach merely because it is eaten after a meal. On this specific proposed mechanism of absorption and rotting, ECU’s assessment is partly aligned with the direction of today’s literature: fruit is still digested and absorbed normally when eaten at different points in a meal, and no evidence supports the mechanism that it will “inevitably rot after a meal” PMID: 41300026、PMID: 41837403

However, “does not rot” is not the same as saying that order has no effect on any physiological response. Professor Lin’s original article at the time regarded eating order as a matter of personal habit. The literature has since been updated: subsequent randomized trials and systematic reviews indicate that eating vegetables, fruit, or protein before carbohydrates may reduce postprandial glucose and insulin responses. This metabolic dimension was added as science evolved; it does not rewrite the original article’s historical position as a later update made by the professor himself PMID: 41837403、PMID: 42045803、PMID: 39777075、PMID: 37892527

Therefore, there is no reason for alarm if the sole concern is that fruit eaten after a meal will “rot in the stomach.” For people with diabetes, gestational diabetes, or a need to manage postprandial blood glucose, food order may be a useful detail within comprehensive care, but it cannot replace overall dietary quality or medical advice PMID: 41837403、PMID: 42045803

Dietary patterns matter more than a single slogan

Long-term health is generally not determined by one food or one fixed macronutrient ratio. Patterns represented by different healthy-eating scores have all been associated with lower all-cause mortality in long-term prospective cohorts. This supports tailoring a healthy diet to culture, preference, and sustainability rather than declaring that only one menu can give everyone a long and healthy life PMID: 36622660

Research on the Mediterranean diet also illustrates how to read the evidence correctly: greater adherence is associated with a lower risk of all-cause mortality, but observational cohort studies may still be affected by self-reported data and residual confounding. An “association” cannot be translated into a guarantee that anyone who follows the diet will live longer PMID: 38819819

Likewise, the proportion of carbohydrates alone is insufficient to judge the quality of a diet, because the nutrient sources that replace them change the health associations. The prospective human data included by ECU do not support declaring an extremely high-carbohydrate diet the one universally healthiest, longest-lived answer PMID: 30122560

The same food label can conceal very different forms of processing

“Oats,” “fruit and vegetables,” and “plant-based” are not seals of quality. The structural integrity, thickness, and processing of oats affect postprandial glucose and insulin responses. Unsweetened oats also cannot be treated as the same exposure as ready-to-eat cereal with added sugar, flavorings, or other ingredients PMID: 33296453、PMID: 34977959

Frozen fruit and vegetables should not automatically be judged nutritionally inferior. Preprocessing may reduce some water-soluble vitamins, but differences between fresh and frozen products depend on the item, nutrient, storage, and handling. The blanket claim that “frozen can never match fresh” runs counter to the direction of the literature reviewed here PMID: 42255710

For ultra-processed foods, short-term experiments must be distinguished from long-term inferences about disease. A randomized crossover trial found that a diet rich in ultra-processed foods led participants to consume more spontaneously and gain weight in the short term. Long-term observational studies have also identified risk associations that warrant attention, but they do not establish that ultra-processed foods directly and inevitably cause every disease outcome PMID: 31105044、PMID: 41978089、PMID: 41274499

Gluten-free and carnivore diets: first ask who they are for

A gluten-free diet is not wholly pseudoscientific. In people with celiac disease, gluten triggers immune-mediated damage to the small intestine, and a strict gluten-free diet is the principal treatment. Some populations without celiac disease may also benefit, so a categorical dismissal would contradict the literature PMID: 42273246、PMID: 42344820

Conversely, people without a diagnosed gluten-related disorder should not automatically treat “gluten-free” as a synonym for healthier or easier weight loss. The nutritional quality of processed gluten-free products does not point uniformly in one direction; it must be assessed from the actual formulation and the individual diagnosis, not inferred directly from the label PMID: 40778429、PMID: 37285813

Nor should personal testimonials elevate the carnivore diet into a universal treatment. The number of human studies has increased since the original article was published; in that sense, the literature has been updated. Yet the updated scoping review still finds evidence on long-term effectiveness and safety to be very limited and does not support recommending the diet broadly for long-term use at present PMID: 41599961

A carnivore dietary pattern may lead to insufficient fiber and certain micronutrients, and those risks deserve serious consideration. Existing reviews, however, generally describe them as “potential risks” and also record a small number of favorable short-term indicators. It would therefore be inaccurate to write that every adverse outcome will inevitably occur in every person following the diet PMID: 39796574、PMID: 41599961

Tests, interactions, and the language of risk

A positive food-specific IgG4 result generally reflects repeated exposure and immune tolerance; it should not be used to diagnose food allergy or intolerance. This does not mean that all allergy blood tests are useless. In the appropriate context, a professionally interpreted specific IgE test can form part of a diagnostic evaluation PMID: 18489614、PMID: 22835332、PMID: 42111357

Nor can risk be ruled out by the experience of friends or relatives. The clinical significance of grapefruit–drug interactions varies with the medication, individual differences, product warnings, and patient vulnerability. Although it cannot be predicted uniformly, useful clinical evidence and management pathways do exist. People taking medication should follow the warnings on their prescription packaging and consult a physician or pharmacist PMID: 27709656、PMID: 23184849

How to read diet news

When you see phrases such as “you must,” “one hundred percent,” “extremely toxic,” or “suitable for everyone,” first ask who was studied, whether the research concerned a food or an extract, whether the outcome was a short-term physiological marker or a disease, and whether the design was a randomized trial or an observational study. Data on processed meat and colorectal cancer support a modest risk association but are also subject to confounding. It is therefore wrong both to claim there is no risk and to replace risk communication with the language of acute poisoning PMID: 32029911、PMID: 40210826

The most robust conclusions in nutrition usually come with boundaries: acknowledge what is known, identify what remains unknown, avoid presenting association as causation, and allow assessments to change when the literature is updated. This approach preserves Professor Lin’s original text and historical position while enabling readers to make more precise choices using today’s evidence PMID: 36622660、PMID: 38819819