What is new? A 2026 review compared keto with the Mediterranean diet
On September 21, 2026, Xu and colleagues published a systematic review and meta-analysis titled “Comparative effects of ketogenic diets and mediterranean diets on obesity and metabolic outcomes.” It included randomized trials in adults and searched the literature through March 2026. Compared with Mediterranean diets, ketogenic diets were associated with greater short-term reductions in body weight, body mass index, waist circumference, body-fat percentage, and insulin. However, there was no clear difference in fasting glucose, HbA1c, LDL or HDL cholesterol, triglycerides, blood pressure, or C-reactive protein. Publication bias was detected for the LDL and triglyceride outcomes, so those estimates require caution [1].
What changed is that we now have a newer direct comparison between two widely used patterns. What did not change is equally important: the review did not prove that keto is better for every person, that its weight difference persists for years, or that it prevents cardiovascular events or diabetes complications. The authors framed keto's advantage as short term, while describing the Mediterranean pattern as a potentially more sustainable strategy for long-term cardiometabolic health [1].
Another recent review asked a narrower question: when prescribed energy intake is comparable, does keto produce more weight loss? Its meta-analysis contained only six trials, 259 participants, and follow-up from 6 to 52 weeks. The average difference was small and favored keto, but certainty was low because the studies were small and affected by risk of bias and imprecision. When the analysis was limited to studies with comparable prescribed protein, the estimate became inconclusive. The practical conclusion is that a short-term difference may exist, but the evidence does not show that ketosis itself makes body fat disappear independently of energy intake, protein intake, and adherence [2].
Nutritional ketosis is not diabetic ketoacidosis
A ketogenic diet is very low in carbohydrate, relatively high in fat, and adequate in protein. When carbohydrate intake falls substantially, the liver produces ketone bodies from fat and the body uses some of them as fuel. This is nutritional ketosis. Not every lower-carbohydrate diet produces ketosis, and a limited presence of ketones does not automatically mean an emergency [3, 13].
Diabetic ketoacidosis, or DKA, is different. Too little effective insulin allows ketones and acids to accumulate enough to disturb the blood's chemical balance, often with severe dehydration. It is a medical emergency, most often associated with type 1 diabetes but possible in other settings. Warning symptoms include marked thirst and urination, vomiting, abdominal pain, deep or rapid breathing, fruity-smelling breath, drowsiness, or confusion. Do not try to treat this by changing food or fluids at home; seek emergency care immediately [7, 8].
This distinction is especially important for people taking SGLT2 inhibitors used in some diabetes, heart, and kidney conditions. The US Food and Drug Administration warns that ketoacidosis with these medicines can occur even when blood glucose is not as high as expected. Severe calorie or food restriction, illness, dehydration, surgery, insulin reduction, and pancreatic insulin deficiency can increase risk. Do not start keto or reduce insulin or another diabetes medicine on your own. If you have diabetes and develop severe nausea, vomiting, abdominal pain, breathlessness, or unusual fatigue, you need urgent assessment even if the glucose reading is not high [7, 9]. For a broader explanation that does not confuse insulin resistance with a diabetes diagnosis, read what insulin resistance means.
Potential benefits and evidence limits: short term is not the final verdict
Before the newest reviews, the broadest high-level synthesis was a 2023 umbrella review that combined 17 meta-analyses and 68 randomized trials. Follow-up was usually short, with a median of 13 weeks. It found moderate-quality evidence for reductions in body weight and HbA1c, and high-quality evidence for lower triglycerides in some comparisons. In contrast, there was high-quality evidence for an increase in LDL cholesterol with some ketogenic patterns and moderate-quality evidence for increased total cholesterol. Many other associations were supported by low or very-low certainty evidence [3].
This means a person with excess weight or type 2 diabetes may see improvements in weight or selected glucose and lipid markers during follow-up. It does not mean keto cures insulin resistance by itself or that everyone responds in the same way. A rapid carbohydrate reduction can lower some people's glucose-medication requirements, which may be useful under supervision but also creates a hypoglycemia risk if medicines are not reviewed. Guidance reviews emphasize that LDL responses to very-low-carbohydrate diets are mixed and that evidence on effectiveness and safety beyond two years is sparse [6, 9].
The word “keto” also covers substantially different interventions: a therapeutic epilepsy diet delivered by a specialist team, a medically managed very-low-energy ketogenic program, and an informal diet assembled from ordinary or packaged foods. Evidence from one cannot simply be transferred to all the others. Early loss of glycogen and associated water can also move the scale quickly without representing an equal amount of fat loss. Real-world value depends on adherence, food quality, adequate protein, fiber and micronutrients, the person's health, and their medicines [2, 3, 6]. If you are comparing keto with meal-timing strategies, our guide asking whether intermittent fasting is suitable explains why timing is a different tool rather than a metabolic shortcut.
Common effects, cholesterol, and stones: risk is not identical for everyone
During the first weeks, some people experience constipation, nausea, headache, dizziness, fatigue, cramps, or temporarily reduced exercise tolerance. A recent systematic analysis of ketogenic interventions found that gastrointestinal events were the most frequently reported and constipation was the most common individual event. However, it combined different therapeutic diets and patient groups, and adverse-event reporting was not standardized, so its percentages should not be treated as a fixed personal forecast [4].
Constipation and fiber shortfalls are not inevitable, but they become more likely when a plan removes legumes, fruit, whole grains, and vegetables without thoughtful replacements. The same concern applies to folate, potassium, magnesium, and selected vitamins. “Keto-friendly” does not mean nutritionally balanced. A pattern built around butter, processed meat, and cheese differs from one containing varied vegetables, nuts, olive oil, fish, and less-processed protein sources. Fat quality matters. American Heart Association guidance favors unsaturated fats in place of saturated fats, healthy protein sources, minimally processed foods, and a variety of vegetables and fruit [12].
Blood-lipid responses vary. Triglycerides may fall and HDL cholesterol may rise in some people, while LDL cholesterol rises substantially in others. Weight loss does not remove the need to review laboratory results, family history, and overall cardiovascular risk [3, 6]. For the kidneys, it is inaccurate to say keto damages every healthy kidney, but stones are a documented risk in ketogenic-diet use. A meta-analysis of 36 studies involving 2,795 people estimated a 5.9% pooled incidence over an average follow-up of 3.7 years, with substantial variation in ages, indications, and diet types. Previous stones, kidney disease, or dehydration changes the decision and is not a reason to self-prescribe supplements or salts [5, 11]. If you have stone symptoms, read when a kidney stone may pass at home and when it needs medical care.
Who should not start keto without clinical supervision?
If you have type 1 diabetes, use insulin, take a sulfonylurea, or take an SGLT2 inhibitor, do not start severe carbohydrate restriction on your own. Insulin and some tablets can cause hypoglycemia when food intake changes rapidly, while SGLT2 inhibitors are associated with ketoacidosis that may occur without markedly elevated glucose. Medication adjustment belongs with the prescribing clinician and structured monitoring. Do not stop or reduce a medicine based on an article or a single reading [7, 9].
Pregnancy and breastfeeding are not appropriate times to experiment with keto for weight loss without the clinicians caring for the mother. A systematic review of carbohydrate restriction during lactation found that the evidence consisted almost entirely of case reports and small case series, and many reported mothers were hospitalized with vomiting, weakness, nausea, abdominal pain, malaise, and fatigue. This does not prove that every reduction in carbohydrate causes acidosis, but it shows that safety is uncertain and that severe restriction alongside lactation's demands can be hazardous [10]. For questions about weight medicines during pregnancy and lactation, read our guide to weight-loss medicines in pregnancy and breastfeeding rather than mixing the two issues.
Specialist assessment is also important with kidney disease or previous stones, advanced liver disease or liver failure, acute pancreatitis or a relevant history, inherited lipid disorders, cardiovascular disease, underweight, or a current or previous eating disorder. Several rare inherited disorders of fat metabolism are absolute contraindications to therapeutic ketogenic diets. International recommendations for adult ketogenic therapies distinguish absolute contraindications from conditions requiring individualized benefit-risk assessment. The list should not become one blanket judgment about every person or every mild condition [11, 13]. Fatty liver also needs its own diagnosis, fibrosis assessment, and follow-up plan in our fatty-liver treatment and follow-up guide; keto is not a substitute for that assessment, and not every liver condition leads to the same decision.
Food quality and sustainability matter more than a “keto” label
Weight loss is not a moral test of discipline, and keto is not the only way to create an energy deficit. If the pattern creates constant food preoccupation, disrupts social life, or reactivates binge-restrict cycles, that consequence matters and is not a failure of willpower. Reviews do not establish consistent long-term superiority for very-low-carbohydrate diets, and restrictions make adherence difficult for many people [2, 6].
Weight does not inevitably return when ketones disappear. The scale may rise partly as glycogen and associated water are restored, which is not all body fat. Longer-term regain depends on average intake, activity, sleep, medicines, illness, environment, and the transition. Abruptly resuming previous habits without a maintenance plan can lead to regain after many diets, not keto alone. A planned transition to a sustainable pattern is more useful than chasing ketones [2, 6].
A less restrictive alternative is a Mediterranean-style pattern with an individually appropriate energy deficit: vegetables, legumes, whole fruit, whole grains where suitable, varied proteins, and unsaturated fats, while reducing sugary drinks, processed meat, and ultra-processed foods. It is a flexible framework adjusted to laboratory results, culture, budget, and preferences, not a universal prescription or permission to change diabetes medicine [1, 12]. “Electrolyte” products or fat-burning supplements are not an automatic safety plan. Our guide explaining that supplements are used for a clear target, not by default shows why need should be established before choosing a product [12, 13].
What this means for you: base the decision on your goal and risks
If you are considering keto, start by asking what outcome you want and whether a less restrictive option can achieve it. Then review your history of diabetes, blood pressure, lipids, kidney, liver and pancreatic disease, stones, eating pattern, any eating disorder, pregnancy or breastfeeding, and every medicine and supplement. The measurements and tests needed are not a fixed package for everyone; they depend on medical history, examination, medicines, and the proposed plan [6, 9, 11].
Do not start, stop, or change a medicine on your own. If the treating team agrees on a trial of a very-low-carbohydrate pattern, the plan should define its goal and duration, food quality, sources of fat and protein, vegetables and fiber, monitoring, and a transition or continuation strategy. Success is not the scale alone. It includes adherence, symptoms, blood pressure, glucose where relevant, lipids, and organ-function monitoring when indicated [3, 6, 9].
Seek urgent care for repeated vomiting that prevents fluid intake, fainting, severe dehydration, increasing weakness, severe or persistent abdominal pain, or stone symptoms accompanied by fever or reduced urination. If you have diabetes and develop deep or rapid breathing, fruity-smelling breath, confusion, or marked drowsiness, these may indicate ketoacidosis and require immediate emergency care, even with a glucose reading that is not markedly elevated in users of some medicines [7, 8].
For an individual assessment, you can book a consultation with the medical team at La Rose to discuss nutrition and weight management in Maadi or Fifth Settlement, or online nutrition follow-up when suitable. If diabetes, medicines, or concerning symptoms require internal-medicine assessment, that service is in Maadi. The author and reviewer are content-only contributors and are not bookable clinic doctors. The aim is a safe, sustainable pattern, not ketosis regardless of consequences [1, 6, 11].
