Humans have always used ketone bodies for fuel. Even if you've never "gone keto", you too have used ketone bodies. That's because they're present in low levels after a period of even short fasting, such as nightly sleep. After a typical period of sleep, ketone bodies supply 2–6% of your energy requirements, which bumps up to 30-40% after three days without any food.
The official history of the ketogenic diets started in the 1920s as an epilepsy treatment, although ketone bodies themselves were first discovered in the mid 1800s, in the urine of patients with diabetes. From the mid 1900s onwards, ketogenic diets were used less and less, as medications were used for epilepsy control. They came back onto the scene in the early 2000s, in the form of a "modified Atkins" diet used for intractable seizures.
In modern research studies, you may see keto interchangeably referred to as "ketogenic diet", "low-carb ketogenic diet" (LCKD), or "very low-carb ketogenic diet" (VLCKD). Occasionally, you may run into "targeted ketogenic diet" (TKD, where carb intake is concentrated into the hours around exercise) or "cyclical ketogenic diet" (CKD, where a few days of keto is followed by a day or two of higher carb intake). These are more typically referred to in online discussion of bodybuilding diets, and the occasional animal study.
There are five major ways that ketones can become elevated.
First, and most simply, fasting will lead to increased ketone production as your body's glycogen stores are depleted.
Second, prolonged exercise results in increased ketone production, especially if you're already eating a low-carb diet.
Third, diabetes can result in elevated levels of ketones, not only in insulin-dependent type 1 and type 2 diabetics, but also in non-insulin dependent type 2 diabetics. Even if those who aren't yet diabetic, elevated levels of ketones can predict worse blood sugar control and development of diabetes in the following years.
Fourth, supplementing with ketones themselves (i.e. exogenous ketones) or consuming large amounts of coconut fat (which contains medium chain triglyerides that can be broken down into ketones), will raise ketone levels even when not eating low-carb.
Fifth, typical ketogenic diets will cause sustained ketone production due to low carb intake. Typical ketogenic diets can be supplemented with exogenous ketones as well, if higher ketone levels are desired, and MCT ketogenic diets with higher carb intake have been commonly used for epilepsy.
Some food groups are not feasible to include in a ketogenic diet on a regular basis due to carb content, such as grains. Other groups can be eaten ad libitum, such as animal products (other than honey). Many types of foods fall in a middle ground though, such as fruits and dairy. Too much of high-carb varieties can cause you to eclipse your carb limit and be kicked out of ketosis.
Keto is only effective if you maintain it; occasional keto is fine but yo-yo dieting with keto followed by carb binges is not.
Keto dieters tend to eat less than higher carb dieters when given free reign to choose their food. This is a boon for adherence. When looking at all trials measuring adherence, keto had a similar adherence rate as other diets, at around 24%. That being said, carb intake, on average, creeps up over time in longer-term studies, with longer term studies having similar adherence as longer term non-keto studies. This masks individual differences though - some stick to it strictly, others go way off track.
In a two-year open-label (non-randomized) study, dropout rates were 35% for the keto group and 28% for the control group.
The most reliable data for side effects and adverse events come from large trials in children with epilepsy. These show varying rates of a variety of possible side effects, from infrequent gallstones and elevated liver enzymes to more frequent elevated lipids and gastrointestinal upset. Some side effects such as nausea and vomiting may be more likely in the first weeks of the diet, as you adapt to higher ketone levels. Also, acetone can be exhaled and is a fairly accurate marker of ketosis, but can unfortunately also be perceived as "bad breath" by some.
Certain electrolytes (specifically calcium, magnesium, sodium, and potassium) can be low on keto diets. If you're low on any of these, supplementation or changing your food choices to more nutrient-dense foods can reduce side effects.
There are several health conditions for which keto is absolutely contraindicated, such as Porphyria and genetic deficiencies like fatty acid beta-oxidation defects. Other contexts haven't had much if any keto research, and caution is warranted, such as with pregnancy and eating disorders.
For those undergoing a keto diet for serious medical conditions such as epilepsy, accidentally ingesting too much carbohydrate via medication or supplement can disrupt ketosis and indirectly cause danger through reduced diet efficacy. Chewable and liquid medication may have relatively high carb levels, not all of which are accurately labeled.
Several objections have been raised to low-carb diets including keto, positing theories of harm but not yet showing it through trial evidence.
One such objection is that a lack of carbohydrate could be detrimental to gut and immune health. This doesn't appear to be born out in trials, which rather often show either a neutral or beneficial effect for gut health. However, there is much uncertainty here. Researchers have found that ketogenic diets in children with epilepsy reduces levels of certain beneficial bacteria. Whether this has any health implications, though, is uncertain.
Because keto has also been proposed as a general strategy for weight loss and diabetes prevention, some researchers have recommended a more moderate low-carb diet (100-150 grams of carbs) given the relative lack of long-term safety data for ketogenic diets.
There is a perception that micronutrient intake will be low on ketogenic diets. That is not necessarily the case, as keto diets that are well-formulated (similar to the case for any other diet) can meet all micronutrient requirements. A ketogenic diet that's consistently high in red meat, though, could theoretically contribute to iron overload in those who are genetically susceptible. Too much iron is implicated in Alzheimer’s, heart disease, and colorectal cancer. Nearly one million Americans who have genetic susceptibility due to having hereditary hemochromatosis, a condition where you absorb too much dietary iron.
In most studies, keto diets decrease hunger and desire to eat. The most important factor in this may be swapping refined carbohydrate for protein, as protein is the most satiating macronutrient. Ketone bodies themselves may also reduce appetite, although this is still an open research question.
There are two types of studies looking at fat mass: free-living studies (where you assign one group a keto diet and another non-keto, and track their fat loss over time) and metabolic ward studies (where you consign participants in each group to a metabolic ward, to tightly control their diet). The former can be much longer than the latter, but lack of dietary control can make results less reliable. Free-living studies do, however, reflect real-world adherence and regular life better.
Metabolic ward studies show similar fat loss in keto dieters as non-keto dieters, over the course of 2-4 weeks. This is also the result of a well-designed two month metabolic ward study. Free-living studies tend to show slightly greater fat loss in keto dieters, although results aren't consistent.
Out of studies that paired exercise with keto diets, and compared them to non-keto diets plus exercise, the findings are mixed. While the non-keto groups often appear to show better mass retention, much of that may be due to keto groups losing water weight, which is counted as lean mass in body composition measurements.
Initial weight loss on a keto diet (meaning the first week) is nearly entirely water and glycogen, rather than fat. You simply cannot lose several pounds of fat in one week absent extreme situations.
As you ratchet down carbs, you use up liver glycogen. Since each gram of glycogen is paired with about 2.4 grams of water, a full male liver is heavier by 289–432 grams (0.6–1.0 lb), and a full female liver is heavier by 241–364 grams (0.5–0.8 lb). Eventually, you'll tap more and more into your muscle glycogen and empty those stores as well, to the tune of 1.1-1.4 kg (2.4-3.2 lb)PMID: 60579979. Adding this to your liver glycogen, it's easy to lose at least 2-3 kg (5-7 lb) or more of glycogen and water weight.
The ketogenic diet hasn't been tested in randomized trials for acne, eczema, and other common skin conditions. Researchers have proposed that keto may help skin conditions such as acne and psoriasis, and case studies have been reported, but human trials are currently lacking.
On the flip side, two case studies have been published on increased carbs helping treat Prurigo pigmentosa (a rare inflammatory dermatitis) in two people who had been on keto diets. Ketosis was proposed to be involved in the initiation of the condition.