The Evidence-Based Truth About Diabetes: Where the Evidence Ends for Diets, Drugs, Supplements, and “Reversal”

Diabetes: What the Evidence Really Supports, from Control and Prevention to Remission

The Bottom Line

Diabetes can be controlled through an individually appropriate combination of diet, physical activity, weight management, and medication when needed. Some suitable patients may even achieve remission, but no single food, supplement, or dietary pattern can guarantee “reversal” or a cure. PMID: 11832527 PMID: 41020034

First, Distinguish Control, Remission, and Cure

Professor Ching-Shun Lin has long cautioned readers against presenting a short-term fall in glucose, discontinuation of medication, or an individual anecdote as proof that diabetes has been cured. That distinction remains important. Subsequent research has shown that, with structured professional support, suitable people with type II diabetes may indeed achieve remission without glucose-lowering medication through substantial weight management. Whether remission persists is closely tied to maintaining the weight loss. This is conditional remission: it cannot be reproduced by everyone, nor does it mean the disease has permanently disappeared. PMID: 29221645 PMID: 30852132

Lifestyle measures should form a key foundation of care for prediabetes. Metformin can delay type II diabetes in people at high risk, but lifestyle intervention had a larger effect in the trial, so medication is not a universal shortcut that replaces diet and exercise. As for whether metformin has been proven to reduce microvascular disease in people with prediabetes, long-term data have not shown an overall difference between groups. PMID: 11832527 PMID: 26377054

Judge Diets by Context, Not Slogans

A ketogenic or low-carbohydrate diet can be a management option for some people with type II diabetes, and later literature now includes studies reporting that remission may be achieved. This is a more precise account made possible by updated literature and the evolution of science. It does not alter Professor Lin’s historical position in the original article against portraying “reversal” as an achievement available to everyone: remission depends on support, adherence, and sustainability, and it may be lost over time. A strict ketogenic diet, in particular, should not be promoted to every healthy person without individual assessment. PMID: 39163706 PMID: 41020034 PMID: 42466787

Intermittent fasting is not magic that operates independently of energy intake. More recent syntheses of randomized trials indicate that, under comparable energy conditions, it is not superior to continuous calorie restriction. It can, however, be an alternative schedule that some people find easier to sustain. Human trials also do not support alarmist claims that “intermittent fasting causes type II diabetes.” PMID: 39732588 PMID: 40008664

There is no single answer for everyone as to whether eating small, frequent meals or fewer, larger meals is better, because trials in different populations and with different endpoints have produced inconsistent findings. In people with prediabetes, eating protein and vegetables before carbohydrate within the same meal may improve postprandial glucose. That result cannot, however, be extended into proof that the practice causes weight or fat loss. PMID: 34816273 PMID: 31489910 PMID: 30101510 PMID: 37892527

Debates over types of sugar also often misplace the emphasis. There is currently insufficient evidence that high-fructose corn syrup necessarily causes metabolic disease more readily than sucrose. The consistent message is to avoid excessive sugar intake, rather than demonizing one sugar while marketing another as a health product. PMID: 41924324 PMID: 41383338

Natural Does Not Mean Therapeutic, and a Signal Is Not Established Efficacy

Bitter melon is the clearest example. Professor Lin rejected the claim that diabetes can be made to disappear with bitter melon or bitter melon extract alone, a position consistent with the earlier pooled results of randomized trials. The available clinical evidence is insufficient to call bitter melon peptides a reliable form of oral insulin. Updated literature has also produced pooled signals that bitter melon products may offer modest glucose improvement. The more accurate statement today is therefore not that they are uniformly ineffective, but that their effects remain insufficient to support claims of cure, replacement of medication, or the powerful benefits promised in product advertising. PMID: 38274207 PMID: 41280283

Oral peptides do face degradation by digestive enzymes and poor intestinal permeability, but the claim that “all peptides must be broken down into free amino acids before they can be absorbed” is too absolute. The existence of oral semaglutide demonstrates that specialized engineering and absorption-enhancing technology can create exceptions. It also shows precisely why an ordinary supplement cannot claim systemic drug effects merely because its label says “peptide.” PMID: 29263649 PMID: 41939723 PMID: 42076118

Human research on American ginseng has reported glucose improvements, but the magnitude is limited and cannot be equated with a cure. Human signals for white mulberry leaf are confined to a particular formulation, and ordinary products have not been shown to produce reliable glucose lowering or weight loss. Although some research on trivalent chromium has found small improvements in glucose measures, the overall clinical significance remains insufficient to support its use as an effective diabetes treatment. PMID: 25265315 PMID: 41228506 PMID: 24635480 PMID: 31577642

In adults with prediabetes who were not selected according to vitamin D deficiency status, vitamin D supplementation did not significantly prevent type II diabetes. In adults with type II diabetes, neither vitamin D nor fish oil has been shown to preserve kidney function. These findings should not be expanded into a claim that no deficient person could benefit in any disease context. Instead, they underscore that supplementation requires a clear clinical rationale. PMID: 31173679 PMID: 31703120 PMID: 41849024

Effective Drugs Still Carry Risks

Semaglutide is a GLP-1 receptor agonist available in oral and subcutaneous formulations and can be used to lower glucose and reduce weight. It is an engineered and validated drug and cannot be invoked to endorse ordinary oral peptide supplements. This drug class is associated with an increased risk of gallbladder and biliary disease, with greater risk when used for weight loss, at higher doses, and for longer periods. The decision to use it should be made jointly by patient and clinician. PMID: 36608818 PMID: 35344001

Data on canagliflozin likewise remind us that a drug’s benefits and harms must be considered together. An increased risk of lower-limb amputation was observed among people with type II diabetes at high cardiovascular risk. Individual risk assessment is especially important for those with a prior amputation, peripheral vascular disease, neuropathy, or diabetic foot problems, although the drug effects have not been entirely consistent across trials. PMID: 28605608 PMID: 32436638

Glucose Readings Need Context

Healthy people without diabetes normally experience glucose fluctuations throughout the day and after meals; a brief rise does not constitute disease. Different continuous glucose monitors may rank the same set of meals differently, and the same person may have inconsistent responses when repeating the same meal. Strong evidence does not currently support widespread use of these devices for disease prevention or personalized nutrition in people without diabetes. PMID: 31127824 PMID: 32766882 PMID: 39755436 PMID: 41157252

Treatment goals likewise should not simply pursue the lowest possible value. The guideline described by Professor Lin advocates individualized targets for most adults with type II diabetes who receive medication. For older people, those with serious illness, or those with limited life expectancy, the priority may be to reduce symptoms of high glucose and avoid harm from hypoglycemia. Tighter control may reduce some complications but can also increase harm, so age, frailty, medication burden, and personal preferences must all inform the decision. PMID: 29507945 PMID: 35614863

Practical Principles for Evaluating Claims

When you encounter claims such as “natural insulin,” “drink this before meals to reverse diabetes,” “everyone should eliminate a particular food group,” or “a single post-meal spike means disease,” first ask whether the research involved human participants, included a control group, measured short-term glucose or long-term clinical outcomes, and applies to you. Reliable diabetes management does not depend on a single miracle product. It brings the evidence, sustainability, safety, and the individual’s health circumstances together.