Does the Keto Diet Work?

By the Strength from the Well Editorial Team · Evidence last reviewed September 2026 · Not reviewed by a licensed clinician (why we say that)

The short answer

Yes, keto produces weight loss, and usually produces it faster than other diets in the first month, though a large share of that early drop is water rather than fat. In randomized trials that run a full year, the difference between a ketogenic or low carbohydrate diet and a comparison diet is usually small and often not statistically significant. Keto works well for some people and poorly for others, and the clearest predictor of who succeeds is not metabolism but whether the person can live with the restriction.

Key points

The first several pounds on keto are largely water released as stored glycogen is used up. That is real weight but not fat.
At six months keto often shows an edge over other diets; by twelve months that edge usually shrinks to little or nothing.
Triglycerides typically fall and HDL typically rises. LDL cholesterol is unpredictable and rises substantially in some people, so it is worth checking rather than assuming.
Blood sugar often improves quickly, which is why anyone on diabetes medication needs a prescriber involved before starting.
Keto is not metabolically magic. Controlled feeding studies have not found a meaningful fat loss advantage when calories and protein are matched.
Most people stop because of the social and practical cost, not because it stopped working.

What a ketogenic diet actually is

A ketogenic diet restricts carbohydrate far enough that the body runs short of its usual fuel and shifts toward burning fat, producing compounds called ketones that the brain and muscles can use instead. In practice this usually means keeping carbohydrate very low, keeping protein moderate and letting fat make up the bulk of intake. That is a much sharper restriction than what most people mean by low carb.

The diet was not invented for weight loss. It was developed almost a century ago as a treatment for hard to control epilepsy in children, and it is still used that way under medical supervision. Its use for weight and metabolic health came much later.

It matters that keto is a specific metabolic state and not just eating fewer carbohydrates, because many people who say they are doing keto are really doing a moderate low carbohydrate diet. That is fine, and often more sustainable, but the results and the claims attached to full ketosis do not automatically transfer.

Why the first two weeks look so dramatic

The body stores carbohydrate as glycogen in muscle and liver, and glycogen holds water with it, roughly three grams of water per gram of glycogen. When carbohydrate intake drops, those stores are drawn down and the water goes with them. Several pounds can leave in the first week this way.

This is genuine weight, and it can genuinely make clothes fit better and reduce swelling. It is not fat. And it comes straight back when carbohydrate does, which is the mechanism behind the crushing experience of gaining four pounds overnight after one pasta dinner. Nobody gained four pounds of fat from one dinner. That is glycogen and water returning.

Understanding this in advance removes a lot of unnecessary despair. It also explains why early keto results are a poor guide to where you will be in six months.

What the trials show at six and twelve months

The pattern across randomized trials is fairly consistent. At three to six months, low carbohydrate and ketogenic diets tend to show greater weight loss than low fat comparison diets. By twelve months the two groups have usually converged, and the remaining difference is often small enough that it could be chance.

One widely cited twelve month trial randomized adults to a healthy low carbohydrate or a healthy low fat diet, with both groups coached toward whole foods and neither given a calorie target. Average weight loss was similar between the groups and the difference was not statistically significant. What stood out was the spread inside each group: some people lost a great deal and some gained, on both diets.

The researchers had expected to find predictors, specifically that people with a particular insulin response or genetic pattern would do better on one diet than the other. Those predictions did not hold up. That is an honest and slightly uncomfortable finding: we can see that keto suits some people dramatically well, and we cannot yet tell you in advance whether you are one of them.

Is there a metabolic advantage

The strongest version of the keto claim is that ketosis changes energy expenditure so that you burn more for the same intake. This has been tested in tightly controlled settings where participants live in a research facility and every gram of food is measured.

Those studies have generally not found a meaningful fat loss advantage for ketogenic eating when calories and protein are held equal. If anything, some found slightly slower fat loss during the transition period. Energy expenditure changes have been small and not large enough to explain the claims made in popular books.

The practical reading is that keto is a very effective appetite tool for certain people rather than a metabolic loophole. Cutting carbohydrate removes most snack food, most sweets, most alcohol mixers and most fast food in one move, and protein and fat are filling. For someone whose intake was driven by constant grazing, that is a large effect. It just does not require a special metabolism to explain it.

What keto does to bloodwork

Triglycerides usually fall, often substantially, and HDL cholesterol usually rises. Blood sugar and insulin often improve, sometimes quickly, and for people with type 2 diabetes the improvement can be large enough that medication has to be reduced to avoid low blood sugar. That is a supervised conversation, not a self-managed one.

LDL cholesterol is where keto becomes genuinely unpredictable. In most people it changes modestly. In a minority it rises sharply, sometimes dramatically, and that group appears to be enriched for lean people. Researchers are still working out what that means for cardiovascular risk in that particular context, and the honest state of the evidence is that it is unresolved.

The reasonable response is not to panic and not to ignore it. Get a lipid panel before starting and again a few months in, and decide with a clinician what the numbers mean for you specifically.

Who tends to do well, and who does not

People who report the best experience on keto tend to share a few things: they find fat and protein satisfying, their eating problem is frequency rather than portion size at meals, they cook, and their social life does not revolve around shared carbohydrate-heavy meals. Many people with insulin resistance report that hunger quiets noticeably, which is a meaningful quality of life change independent of the scale.

People who struggle tend to be endurance athletes doing high intensity work, people who feel genuinely unwell during the transition and never recover, people who find fat unappealing, and people for whom any all-or-nothing rule sets off a restrict-and-binge cycle. That last group should be careful.

There are also people who should not attempt it without medical guidance at all: anyone on insulin or a sulfonylurea, anyone taking an SGLT2 inhibitor for diabetes, people with pancreatic or gallbladder disease, people with certain rare metabolic disorders, and anyone pregnant or breastfeeding.

Get baseline bloodwork, including a lipid panel, before starting
If you take diabetes or blood pressure medication, talk to your prescriber first
Expect a rough first one to two weeks, and expect water weight to return when carbohydrate does
Keep protein adequate and do resistance training, or you will lose more muscle than you need to
Decide in advance what you will do at a restaurant and a holiday meal

So does it work

It works in the sense that matters: many people lose real weight on it and some keep it off. It does not work in the sense the marketing implies, as a metabolic override that makes calories irrelevant.

If you tried keto and it fell apart, the most likely reason is that it collided with your life rather than that you lacked resolve. A less strict low carbohydrate approach captures much of the appetite benefit at a fraction of the social cost, and for a lot of people that trade is worth making.

And if keto is what finally made hunger quiet down for you, that is a legitimate reason to use it. The best diet in the research is still the one you are actually doing in a year.

Common questions

How long does it take to get into ketosis?

Typically a few days of sharply restricted carbohydrate, though it varies with activity and starting stores. Being in ketosis is not the same as losing fat, and measuring ketones tells you about fuel use rather than progress. Many people do well on low carbohydrate eating without ever reaching or confirming ketosis.

What is the keto flu?

It is the cluster of headache, fatigue, irritability, muscle cramps and poor concentration common in the first one to two weeks. Much of it relates to fluid and electrolyte shifts as glycogen and water leave. It usually resolves on its own. If symptoms are severe or persistent, that is worth raising with a clinician rather than pushing through.

Will keto ruin my metabolism?

No more than any other weight loss approach. All meaningful weight loss lowers resting energy use somewhat, and the effect is related to losing weight rather than to cutting carbohydrate specifically. Keeping protein adequate and strength training protects muscle, which is the part of that equation you can influence.

Is keto safe long term?

We do not have good long term randomized data on ketogenic eating maintained for many years in adults for weight, so the honest answer is that it is not fully settled. Shorter trials have not shown alarm signals for most people, with the exception of the unpredictable LDL response. Periodic bloodwork and a clinician who knows you are doing it is the sensible arrangement.

Sources

This is information, not medical advice. Strength from the Well is an independent publisher. Nothing here is a diagnosis, a prescription, or a recommendation to start or stop any treatment. Talk to a clinician who knows your history before you act on anything you read here — including anything you read here that contradicts them.