Ketogenic diet

Short-term weight loss is real, but don't adopt permanent ketosis as a longevity strategy: the one consistent signal from the trials is a rise in LDL cholesterol, and high-fat eating can blunt your ability to burn carbohydrate when you need it.

It builds fat burning — and only fat burning

Fat-burning capacity can rise a lot. Ultra-endurance athletes who had chosen a low-carbohydrate diet for an average of twenty months burned fat at a peak rate roughly 2.3 times that of athletes on a habitual high-carbohydrate diet — 1.54 against 0.67 grams a minute, reached at a higher exercise intensity — and fat supplied 88% of their fuel at moderate effort, against 56%.^[1] The athletes picked their own diets, so this is a comparison of two groups rather than a trial. The study is reported to have been supported by Quest Nutrition and the Atkins Foundation, with authors receiving book royalties.

Metabolic flexibility, though, means switching both ways: burning fat between meals and carbohydrate once it is available. A high-fat diet improves the first half and weakens the second. See Metabolic flexibility for why that switch matters.

The cost: you burn carbohydrate less well

Direct evidence comes from a small crossover trial in seven trained male cyclists. After only five days of high-fat eating — not long-term keto-adaptation — a day of carbohydrate refeeding refilled their muscle sugar stores to match a high-carbohydrate run of the same trial. Even so, they burned about 30% less carbohydrate, and pyruvate dehydrogenase, the enzyme that gates glucose burning, was blunted at rest, during steady exercise and in an all-out sprint.^[2]

A longer trial points the same way at the level of the cell's fuel sensors. Twelve weeks of ketogenic eating lowered the amount of AMPK in muscle — AMPK being the cell's low-fuel alarm, the very switch that fasting is supposed to turn on — and glucose tolerance was worse, not better.^[3]

LDL cholesterol goes up

The cardiovascular concern about permanent ketosis is a human finding, not an extrapolation from animals. Pooling 47 randomised trials in 3,450 adults aged 35–75 with at least one heart-risk factor, ketogenic diets raised LDL (low-density lipoprotein) cholesterol by about 0.43 mmol/L against a low-fat comparison diet, and raised HDL (high-density lipoprotein) cholesterol. Body weight fell modestly; blood-sugar markers barely moved.^[4] For what an LDL rise means for your arteries, see Cholesterol; for the lean, fit people whose LDL rises far more, see Lean-mass hyper-responders.

Two limits apply. What is missing is data on hard outcomes — heart attacks, strokes, deaths — rather than human data, and the pooled certainty for the LDL signal is graded low.^[5] Other pooled analyses of randomised trials found a smaller but consistent LDL rise, alongside lower triglycerides and higher HDL, and stressed the lack of long-term cardiovascular outcome data.^[6]

Liver and heart muscle: no human harm shown so far

Two other worries circulate. Scarring of the heart muscle has been found only in rodents. In fatty liver disease, randomised trials of lower-carbohydrate diets show better weight and insulin resistance, with limited short-term effect on liver fat or liver enzymes — not harm.^[7]

Cycling in and out of ketosis is untested

Cyclical ketosis — moving in and out of it — is often proposed as the safer route to the same benefits. No trial has compared cyclical with continuous ketosis in people, and the two are not obviously equivalent: the AMPK drop above came from continuous ketogenic eating, so there is no evidence that dipping in and out recreates what fasting does.

What to do

1. Don't treat permanent ketosis as a longevity plan. The benefits are short-term, and the one consistent signal against it is the LDL rise.
2. If you use it, it's worth checking your LDL cholesterol once you're on it, rather than assuming the triglyceride and HDL improvements cover it.
3. If you train hard, know that carbohydrate burning can suffer. In a small trial in trained cyclists, five days of high-fat eating was enough to blunt it, even after carbohydrate was restored.
4. For better fuel switching, exercise is the better-supported lever. See Metabolic flexibility.

Further reading

• Volek JS et al. Metabolic characteristics of keto-adapted ultra-endurance runners. Metabolism 2016.^[8]
• Stellingwerff T et al. Decreased PDH activation and glycogenolysis during exercise following fat adaptation with carbohydrate restoration. Am J Physiol Endocrinol Metab 2006.^[9]
• Nikitara K et al. Cardiometabolic trade-offs of carbohydrate- and fat-focused diets: a network meta-analysis of randomised clinical trials. Eur J Nutr 2026.^[10]
• Zhao J et al. Ketogenic diet-induced changes in adult lipid metabolism: a comprehensive systematic review and meta-regression of randomized controlled trials. BMC Cardiovasc Disord 2026.^[11]
• Chu Y et al. Efficacy and safety of lower-carbohydrate dietary patterns for metabolic associated fatty liver disease: evidence from randomized controlled trials with grade analysis. Nutr Metab (Lond) 2026.^[12]
• Hengist A et al. Ketogenic diet but not free-sugar restriction alters glucose tolerance, lipid metabolism, peripheral tissue phenotype, and gut microbiome: RCT. Cell Rep Med 2024.^[13]