Glossary

Ketogenic Diet

Updated April 1, 2026

A ketogenic diet is a very low-carbohydrate, high-fat eating pattern designed to produce nutritional ketosis. It matters because keto can lower appetite for some people, improve day-to-day blood sugar control in some settings, and create a very different tradeoff profile for training, food variety, and lipids than a standard mixed diet. Peter Attia Drive fitness, nutrition, and longevity advice already frames keto as a context tool rather than a default answer. This page explains what ketosis actually is, where keto tends to help, and where it creates friction that the scale alone can hide.

What counts as keto

Keto usually means keeping carbohydrate low enough that the liver produces ketone bodies at a sustained level. In practice, that often means about 20 to 50 grams of digestible carbohydrate per day, moderate protein, and enough fat to cover the remaining energy intake. Nutritional ketosis is commonly defined as blood beta-hydroxybutyrate at or above about 0.5 mmol/L. People vary in how much carbohydrate they can eat and still stay there, though the useful point is that keto is much stricter than ordinary low-carb eating.

That distinction matters because many people say they are doing keto when they are really doing a loose low-carb pattern. Those are different diets. A ketogenic diet asks you to manage glycogen, fiber, electrolyte intake, and meal structure differently from a moderate low-carb plan.

What keto tends to improve

Keto often works through adherence rather than magic metabolism. Very low carbohydrate intake removes many hyper-palatable foods, narrows meal choices, and can reduce hunger for some people. That can lower total calorie intake without deliberate portion counting.

The cardiometabolic picture is mixed and context-dependent. The 2023 umbrella review of randomized clinical trial meta-analyses found that in adults with overweight or obesity, very low-calorie ketogenic diets were associated with improvements in anthropometric and some cardiometabolic outcomes without worsening muscle mass, LDL cholesterol, or total cholesterol in that subgroup.¹ Separate pooled trials in overweight or obese adults with type 2 diabetes also found meaningful reductions in body weight, waist circumference, and fasting blood sugar markers under ketogenic protocols.²

That is enough to say keto can be effective for some people with obesity or type 2 diabetes, especially when the structure improves adherence. It is not enough to say keto is universally superior to other diets when calories, protein, and long-term consistency are matched.

Where keto creates tradeoffs fast

The biggest tradeoff is carbohydrate availability. Hard interval work, high-volume lifting, and repeated long endurance sessions often feel worse when glycogen stays chronically low. Sun and colleagues reviewed ketogenic diets and ketone supplements in endurance runners and found no clear aerobic performance advantage in the available studies.³ Brooks and colleagues then showed that exogenous ketone monoesters or precursors also failed to improve endurance performance significantly in pooled analyses.⁴

That matters because many people try to solve a training problem with a diet that removes the fuel source their training relies on. Keto can still fit a lower-intensity or adherence-driven phase. It is harder to fit when the main goal is repeated high-output performance.

The second tradeoff is food quality drift. Keto can collapse into processed meats, butter coffee, and packaged “keto” snacks very easily. Once that happens, dietary fat quality worsens, fiber intake often drops, and constipation, poor meal satisfaction, or rising LDL become much more likely.

Lipids and why response varies

Keto does not produce one predictable lipid response in every body type. Some people see triglycerides fall and HDL rise. Others see LDL rise sharply, especially when the diet is high in saturated fat, low in fiber, and started in a leaner body.

The strongest warning data come from leaner participants. In Burén and colleagues' randomized controlled feeding trial in healthy young normal-weight women, a four-week ketogenic diet with 4% carbohydrate, 77% fat, and 19% protein increased LDL cholesterol in every participant, with a treatment effect of 1.82 mmol/L.⁵ A later meta-analysis also found that low-carbohydrate diets increased LDL cholesterol in adults with normal but not high body weight, with baseline BMI strongly shaping the response.⁶

The practical lesson is simple. Keto is a diet that deserves lab follow-up, especially if you are lean, have a family history of premature cardiovascular disease, or build the diet around saturated fat-heavy food choices. “I am in ketosis” does not tell you whether the lipid response is acceptable.

Electrolytes, keto flu, and the first two weeks

Early keto side effects often come from fluid and sodium shifts rather than from ketones themselves. Lower insulin reduces renal sodium retention, glycogen falls, and water leaves with it. That is why headaches, fatigue, dizziness, constipation, and a flat training feel show up so often in week one.

This is where electrolyte-balance becomes practical. A person who sharply cuts carbs and keeps sodium low can feel much worse than a person who lowers carbs more deliberately and actively manages fluids, sodium, and meal structure. The phrase “keto flu” sounds trivial. The underlying implementation problem usually is not.

How to tell whether keto is helping or just narrowing your food choices

This is where The Complete Guide to Macronutrients stays relevant even if you go keto. Protein still needs to be adequate. Fat quality still matters. Fiber still matters. A keto label does not erase those rules.

Common mistakes

Keto gets misread as a free pass on calorie density. Fat is still energy-dense, and “keto desserts” can hide a large surplus inside a very small volume of food.

Exogenous ketones create a second layer of confusion because they are not the same thing as a ketogenic diet. Supplemental ketones can raise blood beta-hydroxybutyrate transiently, though they do not recreate the same food pattern, appetite changes, or glycogen state as sustained carbohydrate restriction.⁴

The bigger question is whether keto is being used for the right job. Keto can be a useful adherence tool for some people trying to lose weight or stabilize glucose exposure. It is often a poor fit for people who care most about repeated high-intensity performance, food flexibility, or long-term sustainability in social eating contexts.

The ketogenic diet makes the most sense when it improves adherence, keeps protein adequate, preserves nutrient quality, and survives contact with your actual schedule. If it breaks blood sugar control out of a reactive loop without wrecking training, labs, or food variety, it can be useful. If it turns eating into a narrow saturated-fat-heavy routine with worse workouts and worse lipids, step back and use a broader low-carb pattern instead.