Ketosis: Physiology, Evidence and Trend

Ketones became an identity before most people understood the chemistry

Keto.

Fat adapted.

Ketone fueled.

Carb dependent.

The language suggests two kinds of humans.

That is not how metabolism works.

Ketosis is a state in which ketone-body production rises enough to become metabolically significant.

It occurs during:

fasting;

very low-carbohydrate intake;

prolonged exercise under some conditions;

certain medical diets;

and pathological states.

The presence of ketones does not tell you whether the situation is beneficial or dangerous.

Context does.

Where ketones come from

When carbohydrate availability and insulin signaling are low enough, adipose tissue releases more fatty acids.

The liver oxidizes those fatty acids.

Acetyl-CoA accumulates.

Some acetyl-CoA is converted into ketone bodies:

acetoacetate;

beta-hydroxybutyrate;

acetone.

These circulate to other tissues.

The brain can use increasing amounts of ketones during sustained ketosis.

Muscle and other tissues can oxidize them too.

This is an elegant fuel-sharing system.

A ketogenic diet tries to maintain this state while eating

Most ketogenic diets restrict carbohydrate strongly, provide moderate or tailored protein and supply a large share of energy from fat.

The exact ratio varies.

Clinical ketogenic diets for epilepsy can be highly structured.

Commercial weight-loss “keto” can be far looser.

Two diets can both produce ketones while differing dramatically in:

saturated fat;

fiber;

micronutrients;

food quality;

protein;

total energy.

Again, the metabolic label does not fully describe the diet.

Nutritional ketosis is not diabetic ketoacidosis

This distinction is non-negotiable.

In nutritional ketosis, insulin is low enough to permit ketone production but generally sufficient to restrain uncontrolled ketogenesis and severe hyperglycemia.

In diabetic ketoacidosis, profound absolute or relative insulin deficiency allows runaway ketone production accompanied by metabolic acidosis and major fluid/electrolyte disturbance.

DKA is a medical emergency.

A ketone meter reading is therefore not interpretable without glucose, acid-base context, symptoms, medications and disease state.

SGLT2 inhibitors complicate the old picture

Some diabetes drugs in the SGLT2 inhibitor class can be associated with euglycemic ketoacidosis.

Glucose may not be dramatically elevated.

That is one reason fasting and ketogenic dieting can become clinically sensitive in people taking these drugs.

The old rule:

“DKA means extremely high glucose”

is not always safe enough.

Ketosis has a real medical history

Ketogenic diets were developed as a treatment for epilepsy.

They remain established in selected drug-resistant epilepsy contexts.

This is important because keto is not simply an internet diet trend.

It has legitimate therapeutic applications.

But a diet can be medically useful for one indication without becoming the optimal universal diet for healthy adults.

That inference has to be earned separately.

Weight loss: ketosis can work, but the mechanism is not magic

People can lose substantial weight on ketogenic diets.

Why?

Possible contributors include:

reduced food choice;

higher satiety in some people;

higher protein intake in some versions;

lower spontaneous energy intake;

rapid early water loss as glycogen declines;

adherence to a clear rule structure.

Ketones themselves may influence appetite signaling.

But long-term fat loss still requires sustained energy imbalance.

The diet does not escape thermodynamics.

Early scale changes can exaggerate fat loss

Glycogen is stored with water.

When carbohydrate intake falls, glycogen stores fall.

Associated water is lost.

This can produce rapid early weight reduction.

That is real body-mass change.

It is not equivalent to the same amount of body-fat loss.

A fast initial result can therefore make the diet feel uniquely powerful before long-term comparison has begun.

Diabetes evidence is more mixed than the marketing

Low-carbohydrate and ketogenic approaches can reduce glucose exposure and medication needs in some people with type 2 diabetes.

Some trials report improvements in HbA1c, triglycerides and weight.

But a 2023 systematic review of randomized trials found no clear additional benefit for glycemic control or body weight compared with control diets over periods up to two years, while HDL and triglyceride differences favored keto.

This is not failure.

It is evidence that the advantage may depend on comparator, adherence and population.

Lipids are the major complication in the “keto fixes everything” narrative

A 2024 meta-analysis of 27 randomized trials found average reductions in:

triglycerides;

glucose;

insulin;

weight;

BMI

but increases in:

total cholesterol;

LDL cholesterol;

HDL cholesterol.

The triglyceride/HDL pattern can improve while LDL worsens.

Metabolism rarely moves one marker in one direction.

LDL responses vary dramatically between people

Some people show small changes.

Others show large LDL rises on very-low-carbohydrate high-fat diets.

Diet composition matters.

Saturated fat matters.

Baseline body composition and genetics may matter.

A 2026 meta-regression synthesis again found average triglyceride and HDL improvements alongside modest mean LDL increases and emphasized the absence of long-term cardiovascular outcome certainty.

That makes the responsible conclusion:

a good glucose or triglyceride response does not erase an adverse LDL response.

“My LDL is high but everything else is good” remains an unresolved risk argument

Some keto communities argue that LDL elevation in lean, insulin-sensitive people is benign.

Current evidence does not establish that reassurance.

LDL is causally linked to atherosclerotic cardiovascular disease across broader evidence.

Whether a particular keto-induced phenotype has identical long-term risk has not been directly resolved by decades-long outcome trials.

The absence of direct outcome data is uncertainty.

Not proof of safety.

Not proof of catastrophe.

Food quality still matters inside ketosis

A ketogenic diet can be built from:

olive oil;

nuts;

seeds;

fish;

avocado;

non-starchy vegetables;

or from:

butter;

processed meat;

cheese;

ultra-processed low-carb products.

Both can produce low carbohydrate intake.

Their micronutrient, fiber and fatty-acid profiles differ.

If ketosis is treated as the only target, food quality disappears.

That is an unnecessary reduction.

Exogenous ketones create another distinction

Ketone salts or esters can raise circulating beta-hydroxybutyrate without reproducing the entire hormonal and metabolic state of fasting or a ketogenic diet.

High blood ketones do not prove high fat oxidation.

They can come from what you just consumed.

This is another reason “ketone level” is not a universal score of metabolic health.

Ketosis can reduce flexibility if identity becomes the goal

A person may begin keto as a tool.

Then the ketone reading becomes the identity.

Carbohydrate use is treated as failure.

But healthy metabolism is capable of using carbohydrate.

Glucose oxidation after a meal is not pathology.

The goal should not be permanent allegiance to one substrate.

That leads directly to the next Article.

The strongest conclusion

Ketosis is a normal metabolic adaptation.

Ketogenic diets can be clinically useful and can improve selected metabolic outcomes for some people.

They are not universally superior.

Comparative weight and glycemic advantages are inconsistent.

Lipid responses can be favorable in some dimensions and unfavorable in others.

And nutritional ketosis must never be confused with ketoacidosis.

Continue exploring

Next: Metabolic Flexibility: What the Concept Actually Means

Keto culture often treats maximum fat burning as metabolic freedom.

Research uses a different idea.

Metabolic flexibility means switching fuel use appropriately when supply or demand changes.

Sometimes that means more fat oxidation.

Sometimes it means more glucose oxidation.