I started writing this as a reply on X to Greg Yang, co-founder of xAI, who recently stepped down from xAI to address his debilitating case of Lyme disease.
Ultimately, I decided that a blog post could provide more detail and allow proofreading by credible scientists, so here we are.
Before we dive in, a disclaimer: I’m not a doctor, nor do I play one on the internet. The following is for informational purposes only. Please consult with your doctor and read the warnings at the end.
***
I’ve had Lyme disease twice, and I’m now 100% asymptomatic. The second case in 2014 was incredibly severe, but it was resolved in 4–6 weeks with an unorthodox but defensible approach: strict ketosis.
I have since replicated the results with 4 out of 4 friends who were effectively disabled by Lyme disease. Compared to a lot of complicated or questionable treatments for Lyme, I view ketosis as a simple “first, do no harm” approach, as most drugs have off-target effects, and we’ve scientifically studied ketosis for more than 100 years. Furthermore, ketosis has been a mainstay of human evolution for millennia. Even good old Hippocrates used fasting in antiquity to treat seizures, which was harnessing ketosis.
In this post, I’ll cover some of the science, simple how-to instructions, and near-future tech options that might sidestep or amplify dietary changes.
Let’s begin with my personal case.
I twice contracted Lyme disease and co-infections on Long Island, confirmed with local testing, best-of-class lab testing in NYC, and lastly with specialists at the Stanford Infectious Diseases Clinic. I mention the three separate rounds of testing, as a lot of people are misdiagnosed with Lyme.
Many conditions have similar symptoms to Lyme disease: Long COVID, Fibromyalgia, Chronic Fatigue Syndrome (CFS/ME), Multiple Sclerosis (MS), Rheumatoid Arthritis (RA), and more. Disambiguating takes proper testing from legitimate MDs, in my opinion. There are simply too many charlatans and well-intentioned amateurs running around.
I had no rash in either instance, which is true in 20–30% of cases. Unfortunately, I believed the local folklore of “no rash, no Lyme” and, in the 2014 instance, I waited until symptoms were debilitating: severe joint swelling, slurred speech, forgetting common words and friends’ names, etc. I didn’t seek proper help until my assistant said, “Tim, I’ve seen you tired, I’ve seen you sick, and this is something else. You need to see a doctor.”
So, I did get help. Sadly, after proper diagnosis and courses of antibiotics, which I still believe are important, most symptoms persisted. I operated at ~10% capacity for 9 months and was on the verge of accepting that my mind, body, relationships, business, and more might be handicapped forever. I felt like I had advanced dementia, fatigue often kept me bedridden, and arthritis-like pain wracked my entire body.
What happened at 9 months?
I started brainstorming subtraction. I’d already tried addition: adding drugs, supplements, and all manner of sketchy “alternative” options. If anything, some of them seemed to be generating more problems.
This is how I returned to the ketogenic diet.
Fortunately, I’d used the ketogenic diet in college for various sports experiments, and I decided to test whether or not sidestepping glucose issues and harnessing anti-inflammation (more on these later) would improve things. I knew I could get into clinical ketosis within 3–4 days.
Within a week, all of my cognitive symptoms were gone.
After roughly 4–6 weeks of a strict ketogenic diet (
But there was one rub. I had no satisfying explanation for why it worked.
I had a few plausible theories, sure, but nothing watertight. I knew the short-term effects of ketosis… but a durable fix? How was that possible? And what were the mechanisms of action?
The lights went on in late 2025 when I again interviewed Dr. Dominic D’Agostino, one of the world’s leading researchers and synthesizers of ketones.
Let’s dive into some of what I learned, starting with the biggest missing piece he provided me.
SALVATION THROUGH STARVATION?
Lyme disease spirochetes (Borrelia burgdorferi) are largely dependent on glycolysis for energy production, as they lack a tricarboxylic acid (TCA) cycle and oxidative phosphorylation pathways. In simple terms, the bacteria that cause Lyme disease primarily use carbohydrates for fuel. I had no idea and simply got lucky.
But is it really as simple as starving the bacteria out?
This might not apply to all cases, as such spirochetes can also use alternate fuels like glycerol and pull off all sorts of evolved tricks. LLMs seem to raise an eyebrow at the above theory, but we have at least an N (sample size) of 5 with a 100% success rate. It’s not a Randomized Controlled Trial (RCT), but a lot of compelling science starts with a few interesting case studies.
This glycolysis-dependent piece of the puzzle seems to be critical, but might there be alternate explanations for why keto seems to work? A few possibilities come to mind, and perhaps they synergize to produce the “remission” I and others experienced.
Below are a few leads.
ENERGY PRODUCTION
A ketogenic diet (KD) has a host of fascinating effects on mitochondria, the so-called “powerhouses” of the cell that generate most of your energy (ATP). This was one of my placeholder theories in 2014, as researchers started exploring this terrain seriously in the 1990s.
KDs can help you produce more mitochondria (biogenesis), increase energy production, and arguably improve mitochondrial quality by recycling and replacing them (mitophagy). Net-net, this could be a reboot of your metabolic machinery—you’re remodeling your mitochondria.
Could that address some of the fatigue symptoms of Lyme? Is 1–2 months sufficient to produce these changes and have them stick? Were my later periods of regular fasting—typically three contiguous days of water fasting per quarter—key for reinforcing what the KD kicked off? We don’t really know.
But even if 1–2 months of KD isn’t enough to overhaul your machinery, there are acute energy benefits that could explain my one-week turnaround of cognitive symptoms.
Ketones supply an alternative fuel to glucose, and in some contexts (impaired glucose metabolism) ketones are vastly superior. Perhaps this is partially why, even if not in ketosis, I will take a ketone salt or monoester before recording podcasts: I’m significantly sharper without having to mainline caffeine in the afternoon and sacrifice sleep quality later.
There are documented cases of Alzheimer’s disease (AD) patients who respond incredibly well to exogenous (supplemental) ketones (listen to 2:21 here), and I’ve heard clinicians describe AD patients who normally fail the clock test (drawing a clock) but who can succeed without any difficulty after a single 10–30-gram oral dose of exogenous ketones. Something interesting seems to be happening. Alzheimer’s is sometimes referred to as “type 3 diabetes.” Could a ketogenic diet fix part of the underlying problem, or are ketones simply working around damage (e.g., amyloid plaques and tau tangles)?
I don’t know, but here’s what I do know: every time I cross ~1 mmol/L blood concentration of ketones as measured by any commercial ketone meter (e.g., Keto-Mojo or Precision Xtra), a light switch is flipped and I have extra gears. I’ve seen this repeatedly since my first keto experiments in the mid- and late 1990s. Of course, your personal threshold will likely differ, but I can turn this on at will with exogenous ketones, a 1–2-day water fast, or a 2-4 day low-calorie KD.
Curiously, these extra gears seem to often kick in for many people who are not obviously glucose-impaired. Why? Perhaps it relates to the next bullet.
And perhaps people who aren’t glucose-impaired at the whole-body level (i.e., as assessed by standard blood tests) can nonetheless be glucose-impaired at the brain level? I’d bet money on it.
ANTI-INFLAMMATORY EFFECTS
The primary circulating ketone, beta-hydroxybutyrate (BHB), appears to dial down inflammatory signaling through multiple pathways. In other words, some ketones have direct anti-inflammatory effects.
The next paragraph is technical, so feel free to skip, but there are some juicy bits.
BHB inhibits the NLRP3 inflammasome (a molecular trigger for cytokines like IL-1β), engages the HCA2/GPR109A receptor on immune cells, and may influence gene expression through epigenetic mechanisms like histone β-hydroxybutyrylation. The claim that BHB is a potent HDAC inhibitor is debated in the literature (if true, there are significant cancer implications), but net-net, for some people, ketosis seems to reduce systemic inflammation.
Lyme-induced inflammation can produce vasoconstriction and cause a range of problems (cold hands/feet, dysautonomia), but ketones (D-BHB and L-BHB) have a remarkable effect on improving cardiac output and blood flow by reducing peripheral vascular resistance. This is of personal interest, as both I and my mom have long-standing Raynaud’s Syndrome, but we never thought of any connection to Lyme. My mom has had both Lyme and alpha-gal syndrome, which was transmitted by the Lone Star tick. Now she can’t eat mammalian protein without risk of life-threatening anaphylaxis.
For more on the anti-inflammatory potential of ketones, I suggest you listen to this segment here from my very first conversation with Dominic.
Broadly speaking, anti-inflammatories can have some very fast effects. For instance, if you have minor aches and pains that make it a little painful to walk, you might be able to take two Advil and go for a pain-free walk 60 minutes later.
By extension, how quickly could the anti-inflammatory effects of ketones on a KD translate to less painful or pain-free joints? Smoother and faster cognition? Less or no fatigue? Once again, in my personal experience, all of these and more changed within a week of tipping past 1 mmol/L concentration of BHB as measured by a finger stick with a Precision Xtra or Keto-Mojo device.
Last but not least, guess what? Neuroinflammation impairs glucose metabolism in the brain, and impaired glucose metabolism worsens neuroinflammation.
Everything in this post seems tightly interrelated. That’s good news. Ketosis might be a hammer that hits several nails at once.
METABOLIC PSYCHIATRY
This section will overlap a lot with the preceding two.
Lyme disease is sometimes called “The New Great Imitator” because its symptoms overlap with so many conditions. Some are autoimmune, but many are psychiatric, including but not limited to depression, anxiety disorders, bipolar disorder, obsessive-compulsive disorder (OCD), and ADHD. There’s also emerging data that infection is linked to these disorders and other neurodegenerative diseases.
Dr. Chris Palmer, a Harvard Medical School psychiatrist at McLean Hospital, has developed what he calls “the brain energy theory of mental illness.” In my conversation with him on the podcast, he argued that mental disorders may, in many cases, be metabolic disorders of the brain. His core insight: when brain cells are metabolically compromised, they can become either underactive (shutting down from lack of energy) or hyperexcitable (misfiring when they shouldn’t).
In his model, a ketogenic diet may help by providing an alternative fuel source to struggling brain cells.
The keto–mental health connection isn’t new. Ketogenic diets have been used clinically for 100+ years to treat epilepsy. Centuries ago, churches sometimes locked “possessed” people in a room without food, and lo and behold, the “demons” disappeared after roughly enough time to metabolically switch to ketosis.
Ketogenic diets appear to act through multiple, sometimes overlapping pathways, including those affecting neurotransmission, inflammatory signaling, and gene expression. I bolded those we haven’t directly addressed in this piece.
Furthermore, the ketogenic diet dodges some of the metabolic and off-target side effects associated with many psychiatric medications, especially antipsychotics.
For more on this, I recommend Chris’s book Brain Energy and our full conversation. I also chatted with Dave Baszucki, founder of Roblox, about how he used metabolic psychiatry to save his son, who is diagnosed with bipolar disorder.
HOW TO GET STARTED + FUTURE NON-DIET OPTIONS + WARNINGS
How to Get Started
In brief, if you’ve been diagnosed with Lyme, it might make sense to try 1–2 months of a strict ketogenic diet DURING or AFTER antibiotics but BEFORE you try speculative treatments with non-trivial or unknown downside risks.
Of course, speak with your doctors first.
ChatGPT and similar LLMs can help cover most bases and even meal plans, but be sure to specify “less than 20 grams per day of carbohydrates (CHO).” People can get cute with “net-carbs” and outsmart themselves. I prefer a wide margin of safety when stakes are high (e.g., Lyme symptoms).
If you like books, amidst a sea of terrible options, there are a few that are pretty good.
From Dr. Dominic D’Agostino:
From a CEO I can’t name, who has access to thousands of patients who’ve tried a ketogenic diet in various forms:
My Personal Protocol:
For the most part, I think that trying to eat keto-friendly bagels and faux-desserts is the path to disappointing results.
Especially if you’re just doing a trial run for a few weeks, I like to keep it simple. Do fool-proof first, then, if you want, layer in clever and crunchy after at least two weeks of 1 mmol/L blood readings, and only then with constant ketone monitoring.
To jumpstart keto, I personally like to first do intermittent fasting (IF) for at least a week, only eating within an eight-hour window each day. IF alone can dramatically change your blood work, OGTT, and more. Note that it can take your body 1–2 weeks to overcome the first 12 hours or so of lower energy and occasional irritation, but when you do adapt, it pays long-term dividends.
If you adapt to IF, it’ll make future keto transitions a lot easier and likely eliminate any fogginess, low energy, or “keto flu” symptoms. It helps jumpstart your ketogenic machinery without extended ketosis.
But if you’re in a rush to test keto and want to bite the bullet, you can also just start with a lower-calorie keto diet. The sub-maintenance calories will dramatically speed things up.
In my case, I default to something like the below at a bodyweight of around 175 lbs.:
• 9am Morning – Coffee or tea with 2 tbsp heavy cream (NOT half and half)
• 11am Mid-morning – For the first week of keto and sometimes longer, I’ll mix KetoSpike cocoa exogenous ketone powder into my coffee or tea. This remedies early fatigue.
• Cardio, if any Zone 2 to be done
• ~2pm Lunch – Two cans of chub or jack mackerel mixed with 2 tbsp MCT oil + 2 tbsp apple cider vinegar + salt/pepper. Cheap, fast, and surprisingly good. This will clock in at around 500–700 calories.
• 5pm Weight training, if any
• 7pm or 8pm – Big dinner. This will contain the rest of my calories for the day.
Chopped ribeye on huge salad with extra virgin olive oil, plus a side of creamed spinach
OR
Chicken and cheese plus keto-friendly veggies like broccoli and cauliflower
OR
Lamb chops plus keto-friendly goulash, etc.
You can always add fat with some additional heavy cream in a beverage, as has been done successfully for more than a century with epileptic kids, or a few dollops of sour cream, or a dessert of keto-friendly cheeses.
• Post-dinner – Walk the dogs and curb any glucose/insulin response from the large meal
That’s it. Once you’re in proper ketosis, you probably won’t feel much hunger. It’s quite liberating to reorient to hunger and eating that isn’t compulsive and full of snacking. If you really want to snack, eat more at meals. If you still want to snack, it’s habit and not physiology talking.
Postscript:
Supplements: I take electrolytes (sodium, magnesium, and potassium) as pills or packets at least twice a day. Especially if you’re new to keto, this will be important for avoiding dizziness, cramping, and sleep problems. Just ensure no sweeteners, maltodextrin, etc. are hiding in the product. You’d be surprised what some popular brands do.
Fat: Don’t try to do a low-fat, high-protein version of keto. For reasons we’ll skip here, it’s very hard to make work and not worth the gamble. Aim to consume at least 70% of your daily calories from fat.
Snacks: Be careful with snack foods, even keto-friendly-ish macadamia nuts, which can add up and knock you out of ketosis. Avocados also contain more carbs than you might think. Once mildly knocked out of ketosis, some people need multiple days to regain footing and end up feeling depleted, exhausted, and awful. For beginners, treat ketosis as binary and watch your exact grams of carbs. Play it safe so you don’t end up in metabolic purgatory.
Non-Keto Keto Options and Future Tech
Can you get some of the benefits of a ketogenic diet without eating meat, eggs, and cheese all the time? It sure would be nice.
And, yes, you can make a KD much more appealing, eating a surprising quantity of salads and greens, but I’m always looking for tools and approaches that might make its benefits more accessible.
Here are a few that I’m tracking closely:
Intermittent fasting by itself. At least a 16-hour window of fasting. Read up on neuroscientist Mark Mattson and “flipping the metabolic switch.” Here is one oldie-but-goodie, but note that Mark suggests 16–18 hours of fasting instead of the 12-hour onramp mentioned in that publication. This is also referred to as 16:8 time-restricted eating. 16:8 or 18:6 is a goldmine and perhaps my most surprising personal change of the last two years, in addition to accelerated TMS.
The “metabolic switch” relates to depleting your liver of glycogen, requiring around 16 hours for most people, which then leads to a more ketotic state.
Bioelectronic medicine (e.g., vagus nerve stimulation (VNS)). Dr. Kevin Tracey and others have described the “inflammatory reflex,” whereby vagus signaling can modulate immune activity. Early clinical work has explored VNS in inflammatory conditions (e.g., rheumatoid arthritis). This is not a Lyme treatment per se, but it’s plausible that a VNS device, particularly implants or an ear-based transauricular VNS (taVNS), could be used to decrease inflammation-driven symptoms. There are also some potentially interesting applications to chronic pain management via HMGB1 (special thanks to Ulf Andersson).
Caveat emptor – there is a LOT of BS out there related to vagus nerve stimulation.
Listen to my interview with Kevin, and I’m hoping to help make easier auricular devices more widely available soon. For a possible alternative route, also read up on the applications of famotidine (Pepcid) to the vagus nerve and the inflammatory reflex, which has applications to COVID and more. As always, speak with your doctors before using.
Ultra-low-intensity magnetic approaches (Fareon). Stealth startup Fareon has published preclinical work suggesting microtesla-range magnetic fields can influence neuroinflammation and disease models. This is early science, not clinical guidance, but I became an early investor in this company for a lot of reasons. One of them: I’m hoping it might offer some of the anti-inflammatory benefits of keto with simple at-home hardware. Their tech is not yet available outside of trials, but I’m hoping to help expand that. You can sign up for their email list to be the first to know.
Others?
Do you have other ideas or suggestions? If so, please let me and readers know of this blog post.
Caveats with Lyme Disease
It bears repeating: Many conditions have similar symptoms to Lyme disease, including Long COVID, Fibromyalgia, Chronic Fatigue Syndrome (CFS/ME), Multiple Sclerosis (MS), Rheumatoid Arthritis (RA), and more. Disambiguating takes proper testing from legitimate MDs, in my opinion.
It is also worth noting, however, that data suggest a ketogenic diet might help with symptoms of nearly all of the above (!):
MS → strongest early clinical evidence
Fibromyalgia → promising pilot data
CFS/ME → strong theoretical fit, weak trials
Long COVID → emerging hypothesis
RA → indirect anti-inflammatory benefit
Returning to the caveats, there is another risk lurking behind the label of “Lyme disease.” Some people go shopping for the Lyme diagnosis. If you keep seeing doctors long enough, especially once you venture into “doctors” at the fringe, I promise that you will eventually get a Lyme diagnosis. Some such patients are simply desperate for any explanation and treatment that can provide relief. Others are subconsciously hoping for an external cause for depression and lethargy caused by issues like a rocky marriage, alcoholism, social isolation, etc. It’s a lot easier to take pills or get IVs rather than fundamentally changing the tectonic plates of your life. I get it, and I’ve been there in different contexts.
In the case of Lyme disease, there are entire cottage industries that have popped up to happily take your money for endless treatment that won’t do much.
So, good to be aware and always ask: If I took Lyme off the table, what else might possibly explain this?
Keto Warnings
If you’re on insulin, sulfonylureas, GLP-1 agonists (e.g., Ozempic, Zepbound), or any SGLT2 inhibitor (the “-flozin” drugs), don’t attempt a ketogenic diet without clinician supervision, as carb restriction/fasting can trigger euglycemic ketoacidosis, and medication doses may need rapid adjustment.
In fact, please be sure to always speak with your doctor, m’kay?
Just note that you might need to offer them some reading on the ketogenic diet, as it isn’t a common intervention. This blog post or linked studies and podcasts offer a few starting points.
Also avoid DIY keto if you are pregnant/breastfeeding, have significant kidney/liver/pancreatic disease (including prior pancreatitis), or have a history of eating disorders.
All that said, overall, humans are incredibly well evolved to handle ketosis, especially for the brief periods of time necessary to notice before-and-after changes in the context of Lyme.
One final addendum from Dominic on ALS, at his request:
“FYI, my friend Deanna Tedone was diagnosed with rapidly progressing ALS 17 years ago (given 3 years to live, at most). Her dad, Dr. Vince Tedone, was a world-famous orthopedic surgeon for the Yankees. He came up with the Deanna Protocol, and we proved efficacy in mice: Metabolic Therapy with Deanna Protocol Supplementation Delays Disease Progression and Extends Survival in Amyotrophic Lateral Sclerosis (ALS) Mouse Model.
Deanna tested positive for Lyme disease, and we think this may have been the cause of her ALS. Their foundation is Winning The Fight, and they’re hoping to fund more research on the link between Lyme and neuro diseases.”
For now, that’s all, folks!
Illness and medicine can be squirrely beasts, and I myself have been tempted to give up at times. It can seem like the deck is stacked against you. But sometimes there actually is a simple light of hope at the end of the tunnel.
The ketogenic diet is not a panacea, but its applications beyond weight-loss are compelling. For some, like me and my friends, they can be life-changing.
I sincerely hope this post is helpful.
All the best to you and yours,
Tim Ferriss
Thanks to everyone who proofread this post. Any remaining mistakes are mine. If you spot errors or have corrections, please leave a comment below. Thank you for reading.
