Implications of the circadian nature of ketones.

Ketosis.  Happens during starvation and also by restricting carbohydrates (and protein, to a lesser degree)… might be important for epilepsy and bipolar disorder, too.


Ketostix measure urinary acetoacetate (AcAc) and reflect the degree of ketosis in the blood probably about 2-4 hours ago.  Blood ketone meters measure beta-hydroxybutyrate (bHB) right now.  bHB fluctuates to a greater degree, eg, it plummets after a meal whereas AcAc takes longer to decline.  AcAc/bHB is usually around 1, but increases after a meal (Mori et al., 1990):Ketone body ratio

Conversely, when glucose levels decline and fatty acid oxidation increases, liver redox potential drops which reduces AcAc/bHB.

Galvin et al., 1968
Ketogenic diet-induced ketosis: 90% fat diet for 9 days.  Ketones ~0.3-9.4 mM.
Starvation-exercise ketosis: 36 h fast, then 2.5 h walk.  0.7-6.9 mM.
Exercise-induced ketosis: same as above, sans starvation.  0.2-1.7 mM

Whether it is Jane Plain’s pee, or Jimmy & Freda’s blood, ketosis register strongest at night.  Why?

It’s not measurement error.  From Cameron (2012):

urine acidity

Urinary acidity increases in the evening, which should favor false negatives on the ketostix.  Urine is more dilute in the evenings which also favors lower ketone readings… but this doesn’t happen; confirmed by blood testing.

It occurs in dairy cows, too (Nielsen et al., 2003):

bHB in dairy cows

And in rats fed a ketogenic diet: bHB is elevated after they’ve been eating all night; similar to humans who have been eating keto all day (closed squares, solid line) (De Gasquet 1977):keto rats

PM ketones are likely higher because of the convergence of a few “normal” biological events. 1) Adipose-derived FFAs are not as robust of a delivery mechanism as keto-buffet in stimulating ketogenesis.  2) Exercise is a known ketogenic stimulus.  Daily activity is a weak substitute for “exercise,” but in combination with a ketogenic diet, it’s enough to contribute to the blood ketone pool.  3) Higher levels of sympathetic nervous system activity during the day promotes both lipolysis and ketogenesis.  None of these could single-handedly explain circadian ketosis; rather, each plays a small role explaining the day-long accumulation of ketone bodies in the blood.  When asleep, they cease, fewer ketones are produced, and those present are slowly cleared.  Rinse, wash, repeat.

Knowing this might be important (referring to the circadian aspect, not the “how”).

I.  The ketogenic diet is a bona fide treatment option for certain types of epilepsy.  Some epileptics exhibit more frequent seizures in the morning: temporal lobe epilepsy in these two studies: De Paolis 2013, Cuppens 2009; frontal lobe epilepsy in this study: Pavlova 2012.  There many instances of distinct circadian patterns in seizure frequency… if there is a consensus, I missed it (admittedly, I didn’t look very hard).  The point is there are some epileptics are more clinically responsive to ketogenic diets:
Efficacy of the ketogenic diet: which epilepsies respond? (Thammongkol et al., 2012)
Is the ketogenic diet effective in specific epilepsy syndromes? (Nangia et al., 2012)
Can we predict a favourable response to Ketogenic Diet Therapies for drug-resistant epilepsy? (Schoeler et al., 2013)

Importantly, there appears to be an interaction between the circadian nature of ketones & seizure frequency, and possibly which forms of epilepsy are most responsive to ketosis.  I’m proposing that some epileptic patients on a ketogenic diet experience fewer seizures at night because of the higher ketones.  Alternatively, other patients on a ketogenic diet might not benefit as much because their fits tend to occur in the morning… a speculative corollary of this is that boosting AM ketones would be helpful in this subgroup.

 II.  The ketogenic diet for type II bipolar disorder (Phelps 2013)

This is an n=2 study, but showed a remarkable effect… and duration was 2-3 years (stick this one in your file titled: “ketogenic diet: long-term safety”).


There also appears to be a subgroup of depressives who exhibit strong diurnal fluctuations in mood.  For example, children in this study were hypomanic at night but went to the doctor in the morning when they were more “depressed.”  (Hopefully, shrinks are taking detailed patient histories and interviewing parents, etc., in order to detect this type of thing.)  And it’s not uncommon for other depressives to exhibit greater feelings of melancholia or negative affect in the morning relative to evening… I’m proposing that similar to epilepsy, certain symptoms of bipolar disorder might manifest in the morning because of low ketones; alternatively, who stands most to benefit from a ketogenic diet… and same speculative corollary as above.


Patrick Arnold might just have a new market for his “KetoForce” product.  A spoonful of instant ketosis might put a dent in melancholia.  And seizures.

calories proper

Be Sociable, Share!
  • Jack Kruse

    I think you’ll find those who seize in the morning are those with an altered ASI and salivary melatonin. This tells the clinician, like me their cycles in the brain for blue light, melanopsin, Vitamin D and Vitamin A are all off. These are the people that need mucho ketones. I have lots of clinical experience with this one especially in trauma patients who’s ASI is way out due to catecholamine release.

  • Thomas Hemming Larsen

    Hi Bill,
    Really interesting observations. There has been so much talk about urine/blood ketones and am/pm ketones. This helped to clarify some of the loose ends although I still think ketone production wrt. meal composition and timing is highly individual.

    • I don’t doubt interindividual variability… with duration of ketosis being another big contributor.

  • Galina L.

    One of the symptoms of asthma is being restless at night and a sluggishness on the morning, and many asthmatics report about big improvement on a LC diets.

    • Thanks for the insight. Asthma, another inflammatory condition that benefits from ketosis.

  • Jim Smith

    Very cool blog dude! It seems to take an enormous amount of effort in order to decipher what we should already know instinctively.

    For example, cooking meat over an open fire. Basic instinct is to roast a hot dog and enjoy the fire. We enjoy it and don’t realize how something so primitive could be healthy. We then learn IR builds a structured layer on a bulk water interface. The absorption of IR produces an area of low entropy within the layer creating a negative charge resulting in better vascular flow. Sulfation of cholesterol occurs which improves metabolic function.

    More energy during the day and better sleep at night. Tryptophan, tyrosine, and E. Coli? Dopamine, seratonin/melatonin, and niacin. Nitric oxide and hydrogen sulfide? What is a signaling gas?

    All things we should intuitively know? Yes indeed! A salty, fatty meal and a nice after dinner cigar with a little booze. Imagine that!

    • This. “A salty, fatty meal and a nice after dinner cigar with a little booze.”
      And a toasty fire in my fireplace… sounds like a plan 🙂

  • dave

    Are you proposing that the ketone bodies mediate these beneficial effects? I had it mind that these effects were due to the decreased supply of glucose to neurons? I don’t know, I am asking. I am exercising too but I never thought that ketone bodies were good.

    • Jack Kruse

      This things are coupled coherently in neurons. Nothing runs in series… is all parallel circuits in the brain. That is why it is a quantum computer.

    • yes. It’s hard to dissociate decreased glucose from increased ketones, but the early papers by Veech suggest the effects of ketones go above and beyond that of glucose… eg,

  • This is interesting bill, but generally epilepsy, like bipolar disorder, follows similar risks for greater seizures at night and with prolonged wakefulness (just as in bipolar, mania is increased with wakefulness and mood will elevate further from sleep).

    Depression is greater in morning because sleep is depressant and also because stress hormones are highest at this time.

    Pediatric bipolar is often related to prenatal trauma / complications and more like developmental disorders like autism / ADHD, vs mood disorders… the kids may have specific damage to their circadian rhythm which gives superficial symptoms similar to bipolar disorder, but few of these children have a mood disorder that is in the bipolar spectrum. Bipolar disorder features shifts in mood that last days or weeks. The mood dysregulation of pediatric bipolar is more like dementia with shifts in energy occurring over 1-2 days, suggesting damage to very basic circadian rhymicity.

    It may be melatonin deficiency. People and children who have these wild mood cycles over 1-2 days have been shown in studies to improve by total dark therapy (no exposure to light) OR melatonin supplements.

    BTW The pediatric bipolar diagnosis, and “ultradian bipolar” are highly contested and most consider fraudulent /not real bipolar; likely biologically real but totally unrelated to mood disorders.

    This doesn’t even describe the patients who have simple emotional dysregulation / reactivity behind their “mood cycles” and no real biological shift in energy otherwise. This is very common in autistic children, who will sometimes be abused by doctors and labeled “pediatric bipolar” . They are then given risperdal to make them managable due to their extreme hypersensitivity to environmental stimuli and their neurological inability to control their outbursts/behaviors.

    • “epilepsy, like bipolar disorder, follows similar risks for greater seizures at night”

      I did see a lot of that in the literature. I’m proposing those who are affected more so at night might exhibit a different effect of keto from those who are affected during the day… it could be a long shot, but some px clearly don’t respond.

  • Many depression types have circadian rhythm; even while in a depression lasting weeks or months, mood/energy may be worse in evening or morning. Most common is for mood to be worst in morning and better in evening. Personally I find in depression, dysphoria is worst in evening, energy is best in evening. I experience more numbness/apathy in morning but greater fatigue too.

    Epilepsy, most common is for seizures to be worse the further from sleep and extendend wakefulness. Glutamate, dopamine, and general excitatory of tone of brain, is greatest in evening. This is also the same etiology for bipolar mania and explains why pretty much all “anti manics” make you sleep. Sleep is the best antimanic and the most integral symptom of a bipolar patient shift into mania.

    There was article about how sleep “unjunks” the brain being passed round paleosphere… yea, sleep clears excessive excitatory tone of brain, and this is how it prevents and reduces mania as well as seizures. Energy / brain firing fubar then fail to occur.

    My thought process is for patients experiencing morning seizures, there is likely 2 etiologies.

    1) Extreme sensitivity to cortisol, insulin resistance, thus brain energy metabolism disturbance -> seizures.

    2) Nocturnal hypoglycemia, resulting in waking seizures.

    It is very unusual for seizure patients to worsen with sleep (like bipolar manic patients) and i would almost certainly suspect
    those people are intolerant of fasting hypoglycemia from sleep or hypersensitive to the 6am cortisol surge. Hypoglycemia is a major trigger for seizures, I’ve seen even seizure free patients seize out from hypoglycemia, so I would put all my $$$ into fasting hypoglycemia -> am seizures. Easy way to test this would be to infuse d5-d10 into sleeping epileptic patient with AM seizure. I BET WEEKS PAY NO MORE SEIZURE ~~ 🙂

    • I’ll take you up on that bet! My patient gets an infusion of KetoForce 🙂

  • This seems to suggest children with early morning seizures have an increase in cortical excitability in morning related to their seizures.

    WIth seizure disorders, as well as bipolar/mood disorder, sometimes shifts from baseline are more important than absolute state so it may be the kids seize in morning because simply waking up, with increased brain excitability/acitivity of the waking state, is a relative seizure risk.

    However, the general rule in epilepsy remains that seizure control is greatest with ample sleep , extended wakefulness and sleep deprivation, along with hypoglycemia, are major epileptic risks because of how they disturb brain energetics and increase / accumulate excitatory neurotransmitters in brain.

  • Ilya Beraha

    Let me throw my 2 cents here and some layman’s insight. I’m pretty sure ketogenesis is not an issue on KD (plenty of dietary ketones anyway) and also that prolonged exercise is not a ketogenic but definitely a lipolytic stimulus. Now about the morning clearance: one thing you didn’t account for is the nocturnal thermogenesis by my good ole friend BAT (Brown Adipose Tissue). Indeed night time in humans is the time for a silent reboot of the system, slowed down metabolism and lowered core body temperature (CBT) as result of CBT lowering hormones as adenosine and melatonin. It’s the time when good ole BAT turns on to provide heat for internal organs’ maintenance-heart, kidneys, pancreas. BAT relies entirely and ONLY on fatty acids (lipolysis) for that so no wonder we have ketone clearance in the morning. In that aspect you would probably be excited to know (don’t know if I had mentioned it before) that ketones act as ingested anesthetics: (I have the full-text somewhere) and also that KD is a good for malignant gliomas too:

    And last but not least, all mentioned here therapeutic means: KD, ethanol (acetoacetate) and smoking converge on ONE signaling path – activation of AMPK and co- factors (eHIF-1a, eNOS, did someone say H2S) which is not only an energy gaudge but main antioxidative and anti-inflammatory path in human evolution.

    • Dave

      As far as i know the bat can use glucose too, this is how it was discovered in human adults but it just happens to be activated by catecholamines, in situations where lipolysis is activated.

      • Ilya Beraha

        Nope, it’s totally lipolythic organ, glucose is transformed into LCFA via lipogenesis de novo (DNL), no direct use of glucose.

        • Ilya, I’ve been following what you’ve been saying about this for years now. I like it, but don’t fully grasp. Have you written about it more extensively somewhere? (or can you direct me to a good article or researcher?)

        • Not sure if it’s related, but I did a few experiments gavaging 14C-labeled lipids into mice at 4 degC… and recovering a ridiculously high amount of 14C in BAT (unpublished).

          • dave

            BAT is seriously involved in glucose clearance is rodents at leat …


            UCP1 is activated by free fatty acids for full activty but BAT still takes up and use a lot glucose.

          • in rodents, definitely agree. They’ve got us beat… Am just curious about the quantitative importance of BAT in clearing glucose in humans under “normal” conditions, eg, overnight.

          • Dave

            Probably very limited role in human, mass is too small i think but on a chronic basis, it might still influence body weight. Because of the difference in mass between the two tissues, i am more a believer in the contribution of the wat browning than the bat activity to whole body energy expenditure. Just a guess.

          • “Beiging” …the flavor of the week!
            Exercise works, and we’ll keep discovering new reasons why it works until it can be put in a pill 🙂

            Irisinogen? Myostatinstatin?
            (jk, I’m pretty sure those two have already failed)

          • dave

            Yes, results on irisin from other labs are not encouraging and rumors are that the irisin story is not going to last

            I did not know that myostatin could cause browning. I have seen a picture of a dog with a mutation of the gene recently, it is quite spectacular.

            Independently of how beiging or browning is achieved, I think it’s a good strategy.
            Spiegelman himself says you can’t put exercise in a pill.

            Time will tell.

          • I don’t know about myostatin and browning either; was just referring to “exercise in a pill.”

            Any potential (theoretical) side effects of pharmaceutical-grade browning that you know of? just curious.

          • dave

            I checked after your reply and myostatin mutation causes what browning, therefore there might be a link.
            For the potential side effects, I have been wondering too.

            I suppose you may rise body temperature but to a lesser extent than with uncoupling agents. That would happen no matter how you raise energy expenditure I suppose though.
            The best described browning agents are the beta3 adrenoreceptor agonists and they work really well in rodents but in human I think that in addition to the bat and wat, they also target the heart (not good). beta1 and 2 are even less specific.
            You really need the fatty acids to be oxidized and not released too.

    • Jack Kruse

      ^^^^ EMF 4, Cold Thermogenesis protocol, and Quantum sleep blog all rolled into one. Bill you need to look at the Quantum sleep blog, especially for this blog.

    • Ilya! thanks for swinging by. BAT… I hadn’t thought of that… actually, I didn’t know. Ha! I would’ve predicted heart and brain were two quantitatively more important tissues responsible for clearing ketones during the night.

    • So, you’re recommending bacon, red wine, and cigars?

      I concur.

      • Ilya Beraha

        Def so at least for the winter.

    • Thomas Hemming Larsen

      My bedroom is around 10-12C at this time of year. Is it possible to recruit new BAT at that temperature? As the temperature decreases does the BAT simply use more fat to keep the body at the same temperature?

  • Michael

    These comments by Ilya Beraha are incredibly intriguing!

    Bill, 98th out of 100 “Must-Bookmark Sites on Nutrition Science” – you just beat Alan Aragon and Anthony Colpo who came in at 99 and 100 respectively. LOL

    • I agree… am very pleased when Ilya swings by.

      …and thanks 🙂

  • BullB Man

    Would this favour PM exercises since the higher ketone supply? Or in the case of maximising ketone production (deeper adaptation?) Does it make more sense to “artificially” elevate ketone production by exercising in the morning? I got confused..

    • I think it still makes more sense to exercise in the morning, around the time of food intake. The fuel will be there…