Does coffee raise blood sugar? Conclusion.

Last month, I started an experiment to better understand how different foods and lifestyle decisions impact blood sugar, using a constant-glucose-monitoring device.

For the last few weeks, I’ve been testing whether or not drinking coffee raises my blood-sugar levels.

The answer may seem obvious as coffee contains no sugar, but some people believe coffee can have an effect, and our first coffee experiment indicated just that.

Since then, I’ve repeated the experiment twice. Here’s what I found.

Planning the coffee experiment

I designed the following experiment: I would drink a cup of coffee and measure my blood-sugar levels two hours prior to and after drinking it. Then I would analyze the data to see if drinking coffee seemed to raise my blood-sugar levels.

To increase the reliability of the experiment, I made sure of four things:

1. I would drink the coffee black – nothing would be added to it.
2. I wouldn’t eat or drink anything else, feel stressed, nor do any form of exercise, 2 hours prior to and after drinking the coffee.
3. I would eat ketogenic.
4. I would go to bed and wake up around the same time as I normally do.

It was coffee time.

Drinking coffee

The house was quiet, but I was up, feeling excited about doing another experiment. Everything was ready – beans, scale, grinder, and coffeemaker. The glucose sensor was safely installed in my body – constantly monitoring any change in blood-sugar levels.

Using the above food scale, I put 0.63 ounces (18 grams) of coffee beans into the grinder (about the same amount as for a double espresso). I turned it on. “Grrrrooooooooooooooooouhhh”, it sounded like an elephant had been let loose in the kitchen!

After 20 seconds, the noise finally stopped. I picked up the coffee-bean container, held it to my nose, and inhaled. “Mmmmmmmm, what a delicious smell!”.

I poured the coffee into the Bialetti coffee maker along with 7 oz (2 dl) of boiling water. A few minutes later the coffee was ready. I immediately took a sip. “Oooh, powerful!”.

The coffee didn’t taste as well as it normally does as I hadn’t added any cream – my favorite coffee is the Creamuccino – coffee with steamed cream – but I couldn’t have that this morning as it would affect the experiment. Still, the coffee was nice and a few minutes later there was no more.

Using the constant-glucose-monitoring app I was watching any change in my blood-sugar levels like a hawk. “Ohhhhhh, interesting!”.

Coffee does NOT seem to raise my blood-sugar levels

The graphs below show my blood-glucose levels over a 4-hour period for three separate experiments. The blue line is the blood glucose on a morning I drank coffee. The red line is on a morning without coffee.

Experiment one

The first experiment indicated that coffee may raise my blood sugar. But only slightly:

blood-sugar-coffee-first-experiment

The graph above could suggest that drinking coffee prevents my blood-sugar levels from dropping to the same levels as the morning when I didn’t drink coffee.

Experiment two

The second experiment indicates that coffee does NOT noticeably raise my blood sugar. There’s a lot of variety in my blood-glucose levels on these two days, but to us it seems that coffee does not raise my blood sugar relative to not drinking coffee (the control).

blood-sugar-levels-experiment-with-coffee-and-no-coffee-2

Experiment three

The third experiment also suggests that coffee does not raise my blood sugar much. My blood-sugar levels are quite flat, both on the morning when I drank coffee and the morning I didn’t, indicating that the coffee didn’t impact my blood-sugar levels much.

blood-sugar-levels-experiment-with-coffee-and-no-coffee-3

Conclusion

The chart below shows my average blood-sugar levels based on the three experiments.

blood-sugar-levels-experiment-with-coffee-and-no-coffee-average

To me, this indicates that drinking coffee does not significantly raise my blood-sugar levels.

But, the chart also indicates that my blood-sugar levels on coffee-drinking mornings are on average slightly higher than on mornings I didn’t drink coffee.

However, this difference is not significant and could be due to random variance in blood-sugar levels (see for example how the blood-sugar levels start out higher on coffee-drinking mornings even though I haven’t had any coffee yet).

No significant change in blood-ketone levels

The below chart shows my average blood-ketone levels on coffee-drinking mornings for experiment two and three (regretfully, I didn’t measure ketones for the first experiment).

blood-ketones-levels-experiment-with-coffee

To us, this graph also seems to indicate that coffee does not raise my blood sugar significantly.

If coffee had raised my blood-sugar levels significantly, we would have expected insulin levels to increase as well, causing a subsequent drop in ketone levels. But there was no significant decrease in ketone levels after drinking coffee.

What happens to your blood-sugar levels when you drink coffee?

Please note: This is a n=1 self experiment and my findings may not apply to you. I am a 36-year old insulin-sensitive male, weigh 152 pounds (69 kg), exercise for 10-15 minutes five times a week, with no history of obesity or diabetes.

Although coffee doesn’t seem to raise my blood-sugar levels, the same may not be the case for you.

What happens to your blood sugar when you drink coffee? Let us know in the comments below.

How to test

To test what happens to your blood-sugar levels after drinking coffee, do this:

1. Eat and drink nothing for 2 hours.
2. Measure your blood glucose – this is your baseline.
3. Drink a cup of coffee with nothing added to it (no sugar, no sweeteners, no milk, no cream etc.). Don’t eat or drink anything else for the next two hours, and avoid exercise and stress (these can impact blood sugar).
4. Measure your blood-sugar levels every 30 minutes up to two hours after drinking the coffee (five times in total including the baseline measurement).
5. Evaluate your results. Is there a significant increase or reduction in your blood-sugar levels? *

* It’s normal for blood-sugar levels to fluctuate during the day even when you don’t eat. Therefore, smaller changes – up or down 0.5 mmol/L (10 mg/dl) for example – could be random variation and not caused by drinking coffee.

What do you want us to test next?

We’ve received lots of suggestions for what else to test, for example sugar alcohols, more artificial sweeteners, fasting, alcohol, and resistant starch. What else would you like us to test?

Thanks for suggesting ways to improve these experiments. As per your request, we’ll measure ketones in addition to blood sugar. However, for the experiments we’ve already done but not yet published, ketones data may lack.

Measuring insulin levels is tricker. Do you know of a reliable and simple way to measure it several times per hour? Let us know in the comments below.

Earlier experiments

Are you interested in our earlier tests? Check out the series:

  1. Why You’re Not in Ketosis
  2. How Much Protein Can You Eat in Ketosis?
  3. What to Eat in Ketosis
  4. Do Artificial Sweeteners Affect Blood Sugar?
  5. Does Coffee Raise Blood Sugar? Preliminary Findings

Advanced low-carb topics

Popular now

Low-carb basics

Q&A videos

46 Comments

Top Comments

  1. Chris
    I am an insulin sensitive person with diabetes. I am female. I've had Type 1 diabetes 44 yrs. Coffee, black, DOES increase my blood sugar! Quite significantly. Everyone's body is different.
    Read more →
  2. Erica
    I've been doing intermittent fasting for a year now, skipping breakfast. I drink black coffee (3 cups) every day. I haven't done the specific experiment you did, but I notice that just prior to eating my lunch (anywhere from 11:30am to 1:30pm) my blood sugar is at its lowest (90 or below). I am a T2D (dx 2008) and was very insulin resistant. In the past year I have lost 50 lbs and my A1c went from 12.6 to 5.8. So for me, coffee helps me fast and does not raise my blood sugar.
    Read more →

All Comments

  1. Erica
    I've been doing intermittent fasting for a year now, skipping breakfast. I drink black coffee (3 cups) every day. I haven't done the specific experiment you did, but I notice that just prior to eating my lunch (anywhere from 11:30am to 1:30pm) my blood sugar is at its lowest (90 or below). I am a T2D (dx 2008) and was very insulin resistant. In the past year I have lost 50 lbs and my A1c went from 12.6 to 5.8. So for me, coffee helps me fast and does not raise my blood sugar.
  2. Evinx
    Looks like your blood sugar readings are about 10% higher on coffee mornings. You didn't think it was significant but it was consistent for all 3 experiments. Do you think this is merely coincidence?
    Reply: #32
  3. Bev Anne
    Last night a friend told me that she tried a different brand of coffee and it raised her blood sugar whereas she had not had a problem with coffee before. I don't know if there was any difference besides the brand but that could make a difference.

    I am not a coffee drinker just because I don't like it but I do find these experiments interesting.

  4. Jakub
    There is a flaw in the experiment. In my honones opinion you should not compare your blood glucose levels from different parts of the day (4:50-8:50 to 5:55-9:55) due to liver functions and sleep cycles. I generally agree with the conclusion anyway!
  5. theintelligentomnivore
    Test an apple vs. table sugar. Test raw carrots vs. table sugar. Test potato chips vs. table sugar.
  6. J. Lance De Foa, MD
    The timing of the third trial was much better matched and the difference in glucose much less. No - coffee starts were 78 & 72 at 4:50 AM, and 75 at 6:10 AM as I read the graphs, while the "coffee" FBS, 65-70 minutes later for test 1 & 2, were 90, 92, & 80, all near 6 AM. This is a bit odd, since your body shouldn't "know" whether you will or won't have coffee that day, but the "pre-coffee" FBS are all higher!

    Maybe you need to enter some randomization -flip a coin as whether it is a coffee morning or not. Or do a week with coffee and a week without. That might show the FBS variation is truly variable and independent of the coffee probability.

  7. Lorletha
    Thank you for doing this. For your next experiment I'd like to see you test 2 sweeteners that a lot of low carbers are using lately; Stevia and Erythritol. Supposedly these 2 should have the lowest effect on blood sugar and insulin. I'm curious to see what you find.
  8. Roger Kiwi
    Thats a relief !!
  9. Chris
    I am an insulin sensitive person with diabetes. I am female. I've had Type 1 diabetes 44 yrs. Coffee, black, DOES increase my blood sugar! Quite significantly. Everyone's body is different.
  10. Kim
    I am also a T1D and I use a CGM. I do not notice a rise in blood sugar levels, but I drink dark roast, which has the lowest caffeine levels. It was suggested to me that it could make a difference if I used a lighter roast with higher caffeine levels. The first morning I drank the lighter roast I did see a slight increase in blood sugar levels. However, there are so many variables that come into play that I don't honestly believe that it is a major factor for me. But I will say that in almost 20 years of online message boards I have seen too many people say that they have to cover coffee with insulin to believe that it affects everyone like it does me.
  11. biggus
    I think you mean insulin resistant Chris.
    Reply: #24
  12. Mr_Right
    To me from my experiments using lifestyle libre over two weeks, I found that coffee certainly does influence my blood sugar levels as does a burst of intense exercise.
    without coffee my sugar levels would quite easily reach 3.8mmol but because I was primarily on a ketogenic diet it didn't really matter as my body had enough fuel to keep going as long as i desired
    between meals. drinking coffee seemed to prevent blood sugar from dropping down to 3.8 mmol during a fasting state. not sure how much of this is down to placebo or other factors such as my body having to expend energy etc to process the drink .
  13. Mike
    This study showed that there was a correlation between higher levels of blood sugar levels and the consumption of coffee. The fact that the difference was not statistically significant means that there is a small chance that you would get the same result if you ran the test again (which means we cannot rule out the possibility that there is a real correlation between higher levels of blood sugar level and the consumption of coffee).
    Reply: #18
  14. Jeffrey
    Thank you for the study. I am T2D diagnosed 6/30/2010. I am well controlled with metformin and Lipitor. I would like to see a study like yours with someone like me. Although going 4 hours with only coffee at the 2 hour mark would be almost impossible for me. I would get too hungry. I could probably do ok if the testing began after waking from overnight sleep, drink the coffee, then wait 2 hours. Maybe that is the procedure you followed. Sorry, I did not read the study carefully.
  15. Kevin
    "No coffee" is not a proper control as the act of ingesting liquid may stimulate your digestive system and affect metabolism and hormone secretions. It seems hot water would be a much better placebo or control arm for your home trials.
  16. 1 comment removed
  17. luis enrique galindo valdez
    Caffeine stimulates glycogenolysis
  18. luis enrique galindo valdez
    no de azucar sino de glucosa y es por la glucogenolisis que estimula la degradacion de glucogeno y lo trasnforma en glucosa
  19. CT
    I would like you to test the effect of psyllium fiber in the reduction of blood sugar. Something like eating a 1/2 cup of white rice with and without the fiber ingestion.
  20. Vicky
    I just spent a week taking exogenous ketones from three different companies, sweetened with Stevia. 1. Annoyingly Sweet was the first reaction 2. Threw me out of ketosis was the second reaction Is that possible? I know of no other reason why that would have happened. It was an expensive experiment for sure and will not use what I have left over which is two full containers.
  21. Kelicia
    Please do a study with coffee and stevia vs coffee and other sugar substitutes. I have not tried a coffee that I can stomach without sweetener added. Sad but true!
  22. matthew blackburn
    Do an experiment with diet soda and check both blood sugar and insulin.
  23. Martin
    Hi, I"d be interested to know if you have an insulin response to coffee.
  24. Liberty

    I think you mean insulin resistant Chris.

    Chris is a Type 1 diabetic, not a Type 2. I am also insulin sensitive and I have Type 1 diabetes. T1D's don't have the ability to produce insulin, so we must inject it. Some T1D's have double diabetes and are insulin resistant as well, but a lot of us are insulin sensitive.

    FYI:
    https://www.t1everydaymagic.com/type-1-and-type-2-diabetes-whats-the-...

  25. Cheryl P.
    I think the results may vary depending on the coffee used. I recently discovered that while Peet's lists their 16 oz. Americano as not having any carbs, Starbucks lists the same size Americano as having 3g of carbs. While not significantly different in the big scheme of things, I think that could make a difference for a person who is very insulin resistant.
  26. Marc
    I would like to see an experiment done with fats: full fat cream, butter, olive oil, different cheeses, and coconut oil. I know drinking oil is a bit strange. I would also like to see an experiment on eggs and liver.

    Thanks

  27. Louis-Rene
    Worth nothing that coffee causes salt depletion in your body, when your body lack salt it triggers insuline since insuline helps retain salt in your tissues.
  28. Amanda
    My only problem is you are a non-diabetic, and I assume never diabetic, insulin sensitive person. Your physiology and metabolism are functioning quite well if not optimally. This just isn't the case with diabetics. You could feed an apple to three different diabetics and likely see three different BG profiles. I still find your experiments interesting though and it makes me wish CGM technology was more affordable.
  29. GD
    So now even coffee can be a bad guy too. WTF are you left with on this super restrictive way of eating? Unsustainable in the long term for non-diabetics.
    Reply: #31
  30. Barbara Simoes
    I'd be curious to know about how coffee affects blood pressure. I bet many readers also have high blood pressure and this would be good information. Thanks.
  31. TeeDee
    I guess we all perceive these things differently. When I used to follow the mainstream advice from dieticians/doctors/government, I found it very restrictive to have to eat low-fat, lean meats, low fat cheese and other dairy, restrict eggs and make sure I consumed 5-10 servings of fruit and veg, plus 6-11 servings of whole grains every day. Now, I feel the freedom to eat any cut of meat, fish or poultry I like, plus savory sauces and whatever veg I feel like having. To each their own; you can choose to go back to what you were eating before the "super restrictive way of eating" you're talking about and it's no one's business :)
  32. Anon
    The blood sugar *before* the time he drank coffee was higher on all three mornings he drank coffee. Unless you believe coffee influences the direction of causality (a pretty good trick), that effect over 3 experiments must be something else.

    It could well be nothing. 3 is a small number. If it was still there after 30 I'd agree it must be *something*, but I would then be asking him to record what he ate 2 hours earlier (and whether the anticipation of drinking coffee led him to make different breakfast choices).

  33. Danielle
    Wow! It looks like your BG went very low (40 mg/dl) on the one chart. Is that normal for you? I'm curious b/c my son has T1D and has a CGM and we would treat a significant low like that with dextrose candies.
  34. Louise
    Please do an experiment on alcohol. I know for me, it LOWERS my blood sugar and then often that makes me hungry. I'd also love more on the effects of alcohol on health. And could it's blood sugar lowering effect be beneficial for diabetics?
  35. Hilary
    This is a current topic in my house at the moment as sugar levels do rise with coffee as we have found out by testing in a similar manner to you. Also different brands of coffee spike it higher, for instance I drink decaf and two brands produced totally different readings, kenko decaf spikes higher than Carte Noir decaf. My husband drinks regular caffeinated coffee and will now be limiting his one cup to after a meal as he is type 2 diabetic and finds his bloods are pretty high after coffee. We both drink it black and obviously unsweetened by the way and we also IF eating in a 4 hr window 2 ketogenic meals.
  36. Otto
    Have you perhaps done the same experiment but with cream added to the coffee? Should be interesting to see whether there is any variation or not. It is possible to measure the insulin response in stead of glucose?
  37. Gregory
    Thanks Dr. for the test, but I'ld like see the results of drip or french press coffee vs espresso. Espresso tends to have much less caffeine in a cup because the water is in contact with the grind for less then 2 min., supposedly it's time that leaches the caffeine out and into the cup. but I don't know for sure, just what I hear.
  38. Hilary
    I'm a Type 1 diabetic for 20 years...coffee DOES raise blood sugars, so much so that even though it has 0 carbs I still have to bolus insulin for it.
  39. Christine
    I have been on this way of eating for a year now. I did not have a lot of weight to lose, but I do believe that it is the way to keep me in good health. I LOVE MY COFFEE!! So a few months ago I did the same experiment, dreading the outcome because I promised myself that I WILL quit my coffee it it affects my blood sugar. I tested just like you. The first day I did filter coffee with 1 tablespoon coconut oil. The next day a long espresso - black and the third day decaf espresso - black. All steady as a rock. In fact my blood sugar went down slightly. I tested it hourly for a few hours after I drank and before of course. Also my blood pressure stayed the same. About 110 over 77. (With a slight variation every day but never went higher after the coffee) Humbly grateful for small mercies. I am 61 years old. And I may add that at the end of the experiment my fingers felt like pin cushions!
  40. Elaine Lynn
    Two things: 1). is there a spot on the website I can go to suggest an experiment? 2). I would love to see this exact experiment repeated except with your normal dollop of steamed cream!
  41. Kelly
    I'm with Elaine Lynn..... redo with the steamed cream..... Also with decaf coffee !! Thanks
  42. Rachel
    I absolutely LOVE my coffee, I have strong black coffee with 1-2 tbsp of cream every morning and I LIVE for it. I tried to give it up once before but there was a massive whole in my life and I couldn't cope with the headaches. Right now I am losing weight slowly but steadily still having my one cup per day, so I will stick with that. I have found that it is much better for me to lose weight slowly not feeling deprived then go all out and not be able to stick to it for more than a month. Once I managed to give up coffee for two weeks, so I did not get the headaches anymore, but I yearned for my morning coffee, I missed it sooooo much. I was surprised by that, I thought once the withdrawals were gone I would not miss it. It seems it is deeply ingrained in me. Makes sense though, as I am 36 years old and have been drinking a daily morning coffee for more than 20 years now.
  43. Darrell
    I'm a type 1 diabetic and coffee raises my blood glucose. I bolus for 4g of carbs for 2 cups of coffee.
  44. Me2
    I read some articles that suggest that the body's response to coffee is greatly affected by whether you are drinking coffee every day for years and how much, vs. being an intermittant coffee drinker. It was said that "chronic" coffee drinkers had higher insulin sensitivity and reduced risk for type 2 diabetes, where as intermittant coffee drinkers were more likely to see an increase in glucose levels and lower insulin sensitivity (these results were from 2 different studies).
  45. Tim C
    You have to do a statistical test, such as a t-test, to determine if there is a statistically significant difference in the different datasets.

    This is a great post, but statistical significance is not derived from eye balling data.

  46. Randy Edwards
    This is a very useless and potentially damaging experiment. The question (hypothesis) should be: Coffee affects blood glucose for DIABETICS with less than totally controlled blood sugar.

    Your blood glucose values were all within a normal range that most diabetics would only hope to maintain. The experiment needs to be done with several individuals whose blood sugar is variable and does not always fall within desired ranges below 120, and certainly not by someone whose blood sugar is as consistently low as yours (about 80 or less).

    Therefore, it is dangerous to potentially suggest to someone with diabetes that coffee has no substantial (who cares about statistically significant if the effect is so small) effect on blood glucose

  47. Ralph
    What really matters is how much insulin it takes to keep the blood glucose within such a tight parameter.

Leave a reply

Reply to comment #0 by

Older posts