Low-carbers beware the breathalyzer


A recent article in the International Journal of Obesity should give low-carbers cause for a little alarm.

Here is what happened to a man in Sweden on a low-carb diet:

We report a case of a 59-year-old man, body mass index 26.6 kg/m2, who began a weight reduction program, partly because of knee pains but also because he was a glider pilot where weight is important. He used a Swedish textbook on obesity treatment written by S Rössner together with the commonly used Swedish VLCD [very low calorie diet] Nutrilett (Cederroths, Stockholm, Sweden), 5 packets/day for 3 weeks, which is an approved standard regimen. This treatment resulted in a weight loss of 7 kg.

During dieting, the man discovered that an alcohol ignition interlock device, installed in an official company car, indicated that he had consumed alcohol and the vehicle failed to start. This was confusing because the man was a life-long teetotaller and was therefore both surprised and upset by the result. As he had been supervising private aviation he had access to a second breath-alcohol analyzer, which indicated a simultaneous BAC ranging from 0.01 to 0.02 g/100 ml.

A VLCD diet (very-low-calorie diet, a protein-sparing modified fast) contains mainly protein along with a small amount of carbohydrate and very few calories, usually fewer than 1000 per day. Just about anyone going on one of these diets will soon be in producing ketone bodies at a pretty high rate. But the same goes for a more traditional low-carb diet as well. If carbs are kept at a low level, ketosis will occur. In fact, it’s desired.

Ketone bodies are water-soluble products of fat metabolism. The body has three ways of dealing with ketones: it can burn them for energy (which it does with great success), it can release them in the urine (which is why Ketostix turn purple), and it can vent them through the lungs (ketones can be detected in the breath of one in ketosis).

But, as it turns out, these ketones riding out on expired air can trip the trigger of a breathalyzer or, as the Swedish man discovered, can prevent a car with an alcohol ignition interlock device from starting. Since most people in this country don’t have such devices on their automobiles, the big problem ketone breath can cause would be an arrest for drunk driving.

Before we get go how a breathalyzer reads ketones as alcohol, let’s briefly go over what ketones are, where they come from, and what they do.

Glucose is necessary for our survival, but it doesn’t have to come from our diet. Our bodies have all the metabolic machinery in place to produce all the glucose we need to function just fine without our having to eat a gram of the stuff. Now if we do eat some glucose in the form of sugar and/or starch, the body uses what it needs and stores the rest first as glycogen and then as fat. When we have little or no carbohydrate in our diet, the body reverses the process and breaks down the stored glycogen, which is released as glucose. When the glycogen is gone the body turns to protein as the raw material for glucose production. The conversion of protein to glucose takes place in the liver and is called gluconeogenesis, a word meaning literally the making from scratch, or genesis, of new glucose.

The liver requires energy to convert protein to glucose. Fat provides this energy. The liver burns fat to run the gluconeogenesis machine, but it doesn’t completely burn the fat to carbon dioxide and water, the typical products of the body’s fat combustion process. Instead the liver converts these incompletely burned fats into ketone bodies and releases them into the blood where they are made available for any tissues that might want or need them.

And boy do the tissues want them! The muscles, the heart, and, particularly, the brain greedily lap them up in preference to glucose and/or fat. These tissues lap them up for a reason. Ketones provide a lot of energy very efficiently. The heart, for example, operates about 28 percent more efficiently on ketones than it does on glucose. (I always think of this fact whenever I read Jane Brody and other diverse nutritional idiots rabbit on about ketones being poisons.)

Despite the brain’s and other tissues’ greedy hunger for ketones, they can only consume so much. At some point they get full and stop eating. The liver however doesn’t stop making the ketone bodies because the liver has to continue making the glucose that the body requires. As long as the liver is cranking out glucose, it is cranking out ketones. When the brain, muscles, heart etc. have had their fill, the ketones accumulate in the blood. The body ditches them as mentioned above through the urine and breath. There are other processes that take place to prevent the ketone levels from going too high that are too complex to deal with in this short post, but rest assured that as long as you aren’t a type one diabetic, your ketone levels will never get to a dangerous level.

But if you are righteously following a low-carb diet – especially in the early days – you may produce enough ketones to register on a breathalyzer should you get stopped on suspicion of being drunk.

Here is a diagram taken from the paper showing the metabolic pathways of ketone metabolism.


You can note that one of the pathways for acetone (a ketone) metabolism is the conversion to isopropanol. The ‘ol’ at the end of isopropanol identifies it as an alcohol. The isopropanol in the breath is what the breathalyzer picks up. At this stage of breathalyzer development, the machines can’t differentiate between various alcohols and so registers them all as ethanol, the alcohol in booze.

In the article the man’s breathe registered 0.01 to 0.02 g/dL ml, which is below the 0.08 mg/dL (or ‘point O eight’ as it’s called here) punishable limit for driving a car, but is close to the punishable range in some countries. But, no one knows at this point to what degree of ketosis this particular subject was in to get the 0.02 mg/dL reading. Nor does anyone know if a normal degree of ketosis in a particular person could cause a reading of 0.08 mg/dL or greater, which could get one in a lot of trouble.

Fortunately, the effects of alcohol and ketosis are not additive. But alcohol can convert to ketones. If you drink spirits alone, you could increase your ketone levels, but the alcohol by itself will trigger the breathalyzer to a greater extent.  If you have a sweetened mixed drink or a beer, there may be enough carb to shutdown ketone production.  In which case, your breathalyzer reading will be from the alcohol alone. So if you’ve had a mixed drink or beer then see the blue flashing lights in your rear-view mirrors, you’ll probably be okay. But, if you are following a ketogenic diet and swilling Jameson (a situation I might find myself in), you could be in big trouble. If, on the other hand, you are following a ketogenic diet to the letter while stone sober and see the lights, you still might be in for some trouble. If you should find yourself in that position and have a reading that exceeds 0.08 mg/dL, just tell the officer that you want a blood test for alcohol, which will show a blood level of zero (assuming you really haven’t been drinking), and you should be exonerated.

Or if you’re really worried about it, drink a little glass of orange juice or throw back a couple of mints or other small candy before you drive. As most people who have tried to keep the Ketostix purple know, it doesn’t take much carb to stop ketosis in its tracks.

Please note: I reserve the right to delete comments that are offensive or off-topic.

Leave a Reply

Your email address will not be published. Required fields are marked *

16 thoughts on “Low-carbers beware the breathalyzer

  1. On a no carb diet, will increasing protein consumption (100 grams of protein can be converted to 58 grams carbohydrate w/c I have read without reference) shut down ketosis? My understanding (and your GREAT elucidations on a lot of medical explanations-thank you) is that Acetyl CoA will enter the mitochondria (kreb cycle) if and only if oxaloacetate (made directly from some amino acids and indirectly thru glycerol,some a.a.’s and glucose via pyruvate ) is present. Acetyl CoA will be converted to ketone bodies otherwise. I do realize protein becomes toxic (maybe this is why it’s the most sating macro nutrient, an evolutionary protection) when the liver is overwhelmed and can no longer convert ammonia into urea. Any enlightenment is appreciated.

    Hi busrider– 

    First, the only way protein can be converted to carbohydrate is via gluconeogenesis, so if you have protein that is going to be converted to glucose, the liver is going to make ketones in the process.  If you eat protein, the digestive system breaks it down into the component amino acids, some of which can be used for gluconeogenesis.  The liver doesn’t care where the amino acids come from, i.e., diet or muscle breakdown, it simply converts them to sugar.

    Gluconeogenesis consumes oxaloacetate, which is essential for the entry of Acetyl CoA into the Krebs cycle.  When supplies of oxaloacetate are low, the liver converts the increasing levels of Acetyl CoA into ketone bodies.

    Hope this explanation helps.



  2. Upon more thinking, oxaloacetate is also a precursor to glucose. Hence the body probably will prefer the amino acid et al ->->oxaloacetate->-> glucose pathway first. Only when the body is sated with glucose will the oxaloacetate then be used by the kreb cycle=> end of ketosis.

    Hi busrider–

    You’re right.  Thus the old saying that fat burns in the flame of sugar.  If you don’t have enough carb to provide all the sugar necessary for the cells that can’t survive without it, your liver makes the sugar you need from protein.  The liver requires energy for this feat, which it gets from fat.  But, because oxaloacetate is consumed in the gluconeogenic process, the liver can’t use it to get the Acetyl CoA, which are *sort of) end products of fat metabolism, into the Krebs cycle for complete combustion.  So, if sugar is plentiful, the liver doesn’t need oxaloacetate so it let’s the Acetyl CoA into the Krebs cycle allowing it to ultimately be completely converted to carbon dioxide and water.  If sugar is up, fat burns completely.  If sugar is down, fat is partially burned with the left over converted to ketones.

    Is is all clear as mud now? 

  3. Hi,

    just couple of mints? Really so small dose of carbs shuts down ketosis? Or is it only for limited amount of time? Then what happens when someone is on low carb maintenance phase (~70-100g carbs/day?) – no benefits from ketone bodies?

    Best regards,


    Hi Martin–

    Yep, just a little shuts it down. Why? Because what you probably view as just a little is really a whole lot. If you haven’t read my earlier post on how little glucose there is in a normal blood sugar – a little less than one teaspoon dissolved in the entire blood volume – you should take a look. The amount of sugar in just a couple of mints is about equal to the entire amount of sugar already in the blood. Why would the liver need to make more? It would stop making sugar and start dealing with the excess sugar from the mints.

    The level of ketosis in maintenance is a different issue entirely. As long as the carb intake is less than what the body needs for it’s normal functioning – about 150 grams per day – the liver will have to make up the difference. While it’s making up the difference via gluconeogenesis, it’s pushing out ketones. So, at the point at which you actually consume the 70-100 gms of carbs, ketosis will shut off, but during the rest of the time, the liver is producing them.



  4. This reminds me of when I first went on Atkins about 9 years ago. I am pastor of a small church and I was approached by someone about my breath, and the inferrence was that I was taking a little taste on Sunday mornings or a big taste on Saturday night. Fortunately others around me recognized the dramatic weight loss and were aware of the phenomeon of keto breath and came to my defense. That was the beginning of my practice of dropping a mint or two before talking to people in public. I’m not Catholic so the communion wine defense wouldn’t have worked!

    Hi Charles–

    Thanks for the interesting story.  It’s a good thing you didn’t get the breathalyzer test; that might have made the tongues wag amongst the congregation.



  5. Here are a few numbers that I thought might be interesting on this topic and that I recently read from a metabolism book:

    – There’s a graph showing that ketone bodies reach their max blood concentration at around the 20-25 days of a fast.

    – 60% of the glucose burnt at rest is for the brain; the brain requires about 5-6g of glucose per hour, which is about 220-210g a day.

    -In an intermediary fast, the glucose consumption goes down to about 100-110g a day. ~40g of it comes from lactate, 15-20g from glycerol and the rest, 30-55g, comes from about 75g of amino acids.

    Hi Max–

    These figures sound about right except for the amount of glucose burnt by the brain.  It is in the neighborhood of 5-6 gm per hour, but that totals to about 120-140 gm per day, which is accurate, not 210-220 gm per day as you wrote.  And this amount is when the body has plenty of glucose from the diet – during fasting or low-carb conditions, that amount drops and is supplemented by ketone bodies.



  6. I’m curious as to why ketoacidosis is a problem for diabetics. Is it just because there’s not enough insulin around to drive ketones into the cells?

    And is it really only an issue for Type I diabetes?


    Hi Dave–

    People without type I diabetes can produce insulin.  When ketone levels get high enough, the pancreas releases a little insulin, which stops the production of ketones.  And people who have normally functioning pancreases typically release a little insulin all the time, which keeps the fat in the fat cells unless it is needed.  People with type I diabetes can make no insulin.  They don’t have the ability to shut off ketone production and they don’t have the ability to retain fat in the fat cells.  Fat pours out of the fat cells, the liver does it’s best to deal with it by producing a ton of ketones, the ketones accumulate in the blood, and because there is no insulin to stop production or delivery of fat to the liver ketones increase, acidifying the blood to life-threatening levels.

    It doesn’t really work this way with type II diabetics because, unlike those with type I, type IIs actually produce excess levels of insulin.

    Hope this helps. 


  7. Great answer, thank you. The usual clap-trap about the “danger of ketoacidosis” on a low-carb diet is even more nonsense than I would have guessed.

    Blood sugar gets all the press for Type I diabetes, but it sounds like a broader metabolic malfunction. If high blood sugar doesn’t get you, something else will.

    Hi Dave–

    Clap trap is exactly what it is.  There is no way for one to become ketoacidotic in the diabetic sense of the word unless one is diabetic.  In truth, if you’re in ketosis you are ketoacidotic.  Ketones are an acid.  They are in the blood.  They make the blood a little more acid.  So, therefore, when you’re in ketosis you’re in keto-acid-osis.

    In type I diabetes people get into a severe, life-threatening diabetic ketoacidosis, which is totally a horse of a different color.



  8. Dr Mike : You wrote the best explanation of a complex topic I have ever read. If all teachers had you ability it would be a different country.
    I just lost 70 lbs doing “low fat” then it stopped
    working ( I was eating too much cereal ) so I switched to Atkins and I’m loving it.
    Thank you for a great website.

    Hi Lou–

    I’m glad you enjoy it.



  9. if acetone mimics ETOH and is released via air exchange is it possible to test acetone levels and equivocate for glucose levels thereby producing an equivalent glucose monitoring?

    No, because you won’t have any acetone in your breath unless you’re producing a lot of ketones. And you won’t be producing ketones unless you’re fasting, going low-carb or are a type I diabetic.



  10. I have found this article extremely interesting as I have a son who goes days with out eating and days with out any sleep. In early November he was pulled over for “no front plate” and ended with a DUI…he did indeed drink about 5 hours earlier(2 beers)…but there was no way he should have blown a .123…
    About 2 weeks later…pulled over again..you guessed it..no plate…and once again was administered a breathalyzer…and blew a .08…he swears up and down that he had no alcohol…but he had also not eaten in over 36 hours…(he has a screwy work schedule and the no eating..is simply due to “forgetting and passing out for 10 hours to go back to work…his no eating is simply a “no time and forgot” type of thing…)
    How would one go about “prooving” that his body was indeed in ketoacidosis…at the time I did ask the judge to administer another breath test followed by a blood test…the judge denied the tests…when he was released from jail…(after 60 hours…and eating a small amount) he was given a breathalyzer test via the police department and blew a mere .003…
    I had obtained a hand held breathalyzer and administered the test at home…over a period of 10 days…he blew any where from .00 to .05 at different times…i would administer the test to myself after he would blow high..and the reading would be .00…gave the test to 7 year old and again .00…something doesn’t seem right…any where you can send me to obtain more info on showing that he was indeed in ketoacidosis quite possibly scewing both tests?
    Thank you again!
    Concerned Mom

    I don’t know where to go for more info. The post was about the one article I had read. The easy thing to do is to ask for a blood alcohol test, which measures the alcohol and not the ketones. I think (but I’m not sure) that one can make that request at the time of arrest. Another thing to do would be to keep mints or Lifesavers or other small candies in the car. Sucking on a few of them should while driving should stop ketosis ASAP. The other suggestion would be to get a front plate for the car.



  11. Thanks for the article it was pretty informative. I do have a question however, while one is in ketosis, is alcohol any more dangerous to your organs or to oneself in general than when one is not in ketosis?

    As far as I can tell, it shouldn’t be.

  12. Hi there

    One of my friend went on ATKINS and dr that she was visiting said that her liver was giving in due to her eating too much protien and too little carbs. How is this possible as the was in ketosis all along?

    Thanks for the wonderful explanations and posts.

    I’m thinking of going on Atkins as I have a constant ‘tiredness’ due to a thyriod gland being underproductive. Is it save or should I consult with my gp first?

    I think your friends doctor should maybe take a look in a basic biochemistry or physiology textbook before making remarks about doing one’s liver in by not eating enough carbs.

    It is always preferable to check with your physician before starting a new nutritional regimen.

    Good luck!


  13. If your strength training your muscles need the protein/amino acids to rebuild muscle, NOT for energy. That’s what Carbs are for, energy; yes most people eat way too many, but it’s not healthy to go on an extremly low carb diet. Your body was meant to run on carbs like oatmeal from the canister and other non-processed carbs and foods in general, exception would be fruits (Healthy but not the best carb for weight loss). I went on a low carb diet and my body lost muscle mass which is bad. Yes whenever you lose weight some muscle will be loss, but not as extreme if you eat healthy carbs.

    And what gives you the authority to make these unequivocal pronouncements? Which, BTW, are incorrect.

  14. Paul on 2/9/08 asked and you responded:

    Q. … while one is in ketosis, is alcohol any more dangerous to your organs or to oneself in general than when one is not in ketosis?

    A. As far as I can tell, it shouldn’t be.

    I have to say, Dr. Eades, that I am absolutely thrilled to find some information on carb-control and alcohol consumption. The most I have been able to research is to take the calorie count into consideration and that the body processes ETOH as a priority. I have been in mild ketosis while consuming moderate amounts of alcohol for 10 weeks now and have lost 20 pounds but have been wondering if I am doing damage to my body using ETOH in combination with a carb-controlled diet. I understand that your answer is not necessarily definitive, but thanks for your opinion!

    ETOH actually helps form ketones. At least that’s how I remember it from my medical biochemistry class.

  15. I have a question about the body using protein for its glucose needs. If the body uses 75g of protein to provide the body with glucose, like the earlier post stated, does that mean if I eat only 75g of protein on a 0 carb diet my body will not get the building blocks it needs because all the protein is being used for energy? The reason I ask is because I have been doing an experiment with myself eating 0g of carbs. I have been doing this for close to 2 years but have encountered some problems. I am thinking because I am eating too much protein (eating about 20%- 25%) and not enough fat (75%-80%). The problems I have been having are very similar to hypothyroid only without the weight gain. (loss of electrolytes I need to supplement constantly with salt magnesium and calcium, I have dry hair, dry skin, water retention, brain fog, low sex drive, heart arrhythmias sometimes, and a joint that aches when I don’t supplement magnesium ect.) When I supplement with electrolytes I fell better and alot of the symptoms reside but I want to know what is causing this problem when I eliminate carbs completely. If this is truly the diet humans were meant to run on then why do I feel this way? Why can the Inuit eat this way but I can’t. I am thinking that my body is using the protein as energy and not being used as building blocks which is why I have poor hair and skin as well as many other symptoms. I tried increasing my fat intake to 85%-90% but I got bad diarrhea and nausea. I would also wake up with morning sickness when I consumed this much fat. So if I was able to consume only 70 g would I need to eat more protein for the energy or can protein be used as energy and building blocks?

    It could very well mean that your protein is all being used to maintain blood sugar, but I wouldn’t presume to diagnose you over the internet. If you are having the problems you describe, you need to seek the services of a physician skilled in the use of a low-carb diet in short order.