Why Regular Keto Doesn’t Work for Brain Health (And What Does)

I want to start with a confession.

More than 25 years ago, long before I developed the SKLeTT Protocol, I tried the Atkins diet. I was a physician. I understood nutrition. I was motivated. And I did what the diet told me to do — I ate almost exclusively meat and chicken, zero carbohydrates, no vegetables, no cheating. I was eating ten hamburgers at a sitting. I used urine strips to check my ketones, and when they turned purple, I assumed I was doing everything right.

Two weeks in, I plateaued. My energy was okay — not remarkable, not the mental clarity I later came to know was possible. And then my body simply revolted. I became progressively nauseated, couldn’t face another piece of meat, and within days I had to stop. The diet became physically impossible to continue.

What went wrong? Almost everything — and none of it was visible to me at the time, because I was working without real measurements, without a clinical framework, and without any understanding of what was actually happening in my metabolism.

That experience planted a seed that took two decades to fully grow. Because the honest answer to why I failed on Atkins — and why millions of people fail on “keto” every year — is not that the underlying concept is wrong. The concept is right. The implementation is almost universally inadequate. And when the implementation is inadequate, the brain benefits that the research clearly documents simply do not materialize.

This article is about the gap between what most people call keto and what actually produces measurable, sustained cognitive benefit. Understanding that gap is the first step toward closing it.

What Most People Call “Keto” Is Not What the Research Studies

Here is the most important thing I can tell you, and I want you to read it carefully: the keto diet you read about in magazines and on social media is not the intervention that produces the brain benefits documented in peer-reviewed research.

The popular definition of a ketogenic diet — high fat, moderate protein, carbohydrates under 20 to 30 grams per day — was originally developed as a medical treatment for epilepsy in children. It was subsequently adapted into a weight loss trend, then a wellness trend, then a lifestyle brand. Somewhere in that translation, a critical piece of information got lost: the brain benefits of ketosis are dose-dependent. They don’t happen at any detectable level of ketones. They happen at specific, sustained, therapeutic levels of blood beta-hydroxybutyrate — BHB — that most people following a standard keto diet never come close to achieving.

This is why so many intelligent, motivated people try keto for brain health, feel marginally better for a week or two, then notice no sustained cognitive improvement, and eventually conclude that “keto doesn’t work for me.” In most cases, keto was never actually tried. Low-carb eating was tried. Therapeutic ketosis was not.

The Measurement Problem: You Cannot Manage What You Don’t Measure

When I look back at my Atkins experiment, one of the most fundamental errors I made is something I see replicated by nearly every patient who comes to me having “tried keto before”: I never actually measured my ketone levels with any accuracy.

Urine ketone strips — the purple strips most people use — measure acetoacetate, one of three ketone bodies the body produces. They are a reasonable proxy in the first week or two of ketosis. But here is what almost no one who sells keto products will tell you: after approximately two weeks of sustained ketosis, the body becomes more metabolically efficient. It begins converting acetoacetate to beta-hydroxybutyrate — the primary and most potent ketone body — and reabsorbing it rather than excreting it in urine. The urine strips turn lighter. They may turn nearly negative. And the person using them concludes they have fallen out of ketosis — when in reality, if they have maintained the dietary protocol, they may be in deeper ketosis than ever before. The measurement tool simply stopped working.

This is not a minor inconvenience. It is a fundamental clinical gap. If you are following a ketogenic diet for brain health and you have never confirmed your blood BHB level, you genuinely do not know whether you are in therapeutic ketosis. You may be at 0.3 mmol/L. You may be at 4.0 mmol/L. The dietary approach that achieves 0.3 and the one that achieves 4.0 will look nearly identical from the outside — but their biological effects on the brain are worlds apart.

The Hidden Carb Problem: Why Your “Clean” Keto Diet May Be Failing You

Let me tell you something I discovered through systematic self-experimentation that surprised even me, and that I subsequently confirmed in hundreds of patients through direct blood ketone testing.

In my current clinical practice, patients who eat exclusively animal protein and eggs, with no vegetables whatsoever, consistently achieve measurably higher BHB levels than patients eating the same proteins alongside salads, broccoli, and other “keto-approved” vegetables. When I began testing this systematically with blood monitors, the results were unambiguous: for patients aiming at therapeutic BHB levels above 4.0 mmol/L, vegetable carbohydrates — even in small amounts — are frequently sufficient to prevent that target from being reached.

I tested this on myself, repeatedly. After eating a single leaf of spinach — a food with genuinely negligible listed carbohydrate content — I would find that within two days, if I continued that pattern, I was opening the refrigerator and mentally calculating whether I could get away with a tablespoon of peanut butter. The carbohydrate craving cycle had been subtly reactivated. I repeated this experiment multiple times. The result was consistent.

The reason is neurological as much as metabolic. The brain is hardwired to seek glucose. When carbohydrates are completely eliminated, the brain eventually accepts the shift and begins preferring ketones — but that adaptation requires an uninterrupted transition. A small carbohydrate signal — even from a leaf of spinach — sends a message to the brain’s reward circuitry that glucose is available. The brain responds accordingly. The craving machinery reactivates at a subconscious level.

Standard keto diets not only permit vegetables — they encourage them. This is appropriate for the goals of standard keto. It is incompatible with the goals of therapeutic ketosis.

The Chemical Interference Problem: It’s Not Just About Food

Here is a discovery that took years of clinical observation and direct patient testing to solidify, and that I believe is almost entirely absent from the mainstream conversation about ketogenic diets.

I have tested this systematically. A colleague of mine completed a 7-day water fast — nothing but water for seven consecutive days. I asked if I could check her ketone levels at the end. Her blood BHB was 0.1 mmol/L. She was genuinely shocked. In her mind, she had consumed absolutely nothing for a week. But something she was putting in her body — something she did not consider food — was suppressing ketogenesis almost completely.

I had a patient who had been eating exclusively red meat and water with butter — the Lion Diet — for two full years. Completely zero-carbohydrate, by any reasonable measure. His blood BHB when I tested him was 0.8 mmol/L. The volume of food he was consuming was sufficient to keep his insulin elevated enough to suppress meaningful ketone production.

One critically important and widely overlooked point about water: not all water is equal for therapeutic ketosis. Many commercially bottled waters contain additives — minerals, electrolytes, flavoring agents — that can decrease ketone levels significantly. Most people never read the ingredient labels on their water. In a therapeutic ketosis protocol, the type of water consumed is not a minor detail. It is part of the protocol.

When Your Body Is Fighting Something You Don’t Even Know About

There is one more variable I want to describe — because I have observed it repeatedly in myself and in my patients — and I have not seen it discussed anywhere in the published literature on therapeutic ketosis.

The research explains exactly why this happens. When the immune system detects a viral challenge, it undergoes what scientists call metabolic reprogramming — a massive, urgent shift in how the body allocates its energy resources. Immune cells switch from their normal mode of energy production to a process called aerobic glycolysis — essentially flooding themselves with glucose to generate the rapid energy needed for immune cell proliferation, cytokine secretion, and pathogen clearance. This metabolic takeover has been quantified: an activated immune response consumes an estimated 25 to 30 percent of the body’s entire basal metabolic rate — an energy demand comparable to military training. The body does not politely request that energy. It redirects it — away from processes including ketogenesis.

I recently had a patient who had been holding steadily in the 3.5 to 4.0 mmol/L range. Over a few days, her levels dropped to approximately 2.5 and stayed there. I interrogated her thoroughly — no dietary changes, no new medications, nothing she could identify as different. She did not feel sick. I made a clinical decision she did not initially welcome: I cut her food intake in half, reducing the insulin stimulus sufficiently for the liver to resume full ketone production despite the competing immune demand. The next day, her levels bounced back to 4.5 mmol/L. Within five to six days she was fully back to her normal range. In retrospect, she had almost certainly been mounting a subclinical immune response — her body quietly fighting something she never consciously experienced as illness.

Here is the critical point: without daily monitoring and clinical guidance, those five days of reduced BHB would have felt like “keto not working.” The patient would have had no idea why. She might have concluded she had done something wrong, or that her body “doesn’t respond to ketosis.” Instead, we identified it, corrected it, and she continued her protocol without interruption.

This is what physician guidance provides that no app, book, or online community can replicate: the ability to distinguish between a dietary failure and a physiological event — and to respond accordingly.

The Research That Changed My Understanding: Ketosis Requires a Threshold

A study from Aarhus University in Denmark, published in 2025, crystallizes what I have been observing clinically for decades. Researchers placed participants on a structured ketogenic dietary intervention — verified ketosis, not self-reported low-carb eating — and used advanced imaging to examine what happened inside their brains.

The critical word throughout that description is verified. The participants were in confirmed ketosis — not just eating low-carb and hoping. The brain changes documented required actual therapeutic BHB levels, sustained over time. This is consistent with a formal review examining physiological ketosis and cognitive function, which concluded that cognitive benefits depend on both the level of ketones achieved and the duration of sustained elevation — minimal brain adaptation at low intermittent ketone levels, substantial adaptation at consistent therapeutic levels maintained over weeks.

Stanford’s Metabolic Psychiatry Findings: The Dose Is the Therapy

The work emerging from Stanford University’s metabolic psychiatry program reinforces this picture from a psychiatric angle. A pilot study enrolling patients with serious mental illness — schizophrenia and bipolar disorder — on a physician-supervised therapeutic ketogenic diet showed striking improvements in psychiatric symptoms, metabolic markers, and quality of life. The investigators were careful to specify that these results came from physician-guided, monitored therapeutic ketosis — not a self-administered popular diet.

What this body of work suggests — consistent with everything I have observed across 25 years of practice — is that the brain’s response to ketones is not a simple switch that activates when you reduce carbohydrates. It is a dose-dependent, duration-dependent, precision-dependent biological process. When that threshold is reached and maintained, the brain changes. When it is not reached, it does not — regardless of how faithfully someone believes they are “doing keto.”

The Keto Flu Is a Solvable Problem — Not a Reason to Quit

Almost everyone who has attempted a ketogenic diet has experienced it: the headaches, fatigue, irritability, brain fog, and muscle cramps that arrive during the first week and drive most people to quit.

This experience is real. It is also, in my clinical experience, almost entirely preventable.

The keto flu is not caused by the absence of carbohydrates. It is caused by electrolyte depletion and dehydration. When carbohydrate intake drops, insulin levels fall. When insulin falls, the kidneys begin excreting sodium — and water follows sodium. The rapid fluid shift produces the symptoms patients interpret as the diet “making them sick.” The diet isn’t making them sick. The unmanaged electrolyte and fluid shift is.

The SKLeTT Protocol includes a specific and detailed hydration and electrolyte protocol — including guidance on the type of water used, since most commercially bottled waters contain additives that interfere with ketogenesis. I will not publish the specifics here because the correct protocol is individualized and requires clinical judgment. But I want to be emphatic: with proper hydration management, the keto flu is largely avoidable. The transition is still a challenge — days three through five, as glycogen stores are depleted and ketone production has not yet ramped to therapeutic levels, are difficult. But with the right support, the vast majority of patients move through it. And what they find on the other side makes the transition period feel, in retrospect, like a very small price.

What Actually Works: The Four Differences That Matter

For anyone who has tried regular keto and been disappointed, here is the framework that distinguishes therapeutic ketosis from the popular diet version:

The tool is not the problem. The application has been imprecise. A ketogenic approach that measures, titrates, eliminates hidden carbohydrate and chemical interference, manages the transition, and sustains therapeutic BHB over sufficient time is a fundamentally different clinical intervention from a popular diet trend — even if both are called “keto.”

The Bottom Line

If you tried keto and your brain fog didn’t clear, your memory didn’t sharpen, your energy didn’t stabilize — you didn’t fail keto. In all likelihood, you did what virtually everyone does: you followed a popular version of a low-carbohydrate diet without any measurement of your actual ketone levels, without a clinical protocol for the transition, and without any way of knowing whether hidden variables were keeping you out of the metabolic state you were trying to reach.

I made the same mistake 25 years ago with Atkins. I was eating almost nothing but hamburgers, using urine strips that were giving me meaningless readings by the end of two weeks, and wondering why I was nauseated and plateauing instead of experiencing the transformation I expected.

The transformation is real. The research confirms it. The clinical cases I have witnessed over 25 years make it undeniable. But it requires precision, measurement, physician guidance, and an honest accounting of every variable that determines whether your BHB is truly at a therapeutic level.

That is what the SKLeTT Protocol delivers. Not a diet. A titrated, monitored, clinically guided restoration of the metabolic state in which your brain — and your immune system — were designed to function.

References