This always puzzled me. Not only because it is a nuisance, but also because I cannot understand it.

Whenever I have to take a blood sample to run some tests on it, I am always told not to eat anything for about 12 hours before the test.

That makes no sense to me. If they want to measure the sugar in my blood, for example, well, for sure what I eat affects that, but I'm always eating! If my blood is high on sugar when I eat my normal meals, guess what, that's how I am normally, 24/7, my whole life, so there is indeed a problem. If I fast before the test, obviously my sugar will be low, right? But it doesn't matter, because that is a forced state, that never actually happens (apart from when I'm taking blood samples).

Guess at the end of the test they should say "well, you will probably be fine, as long as you take another blood test every couple of days."

  • 1
    No, it's not obvious that if you fast your blood sugar will be low. Someone with uncontrolled diabetes may very well have a high blood sugar level despite fasting. And if you eat a big plate of pasta before your test, how do you propose that they distinguish between a normal high reading because of your meal and a sugar metabolism problem? Like all scientific tests, it's crucial that confounding factors are controlled.
    – Carey Gregory
    Commented Jul 8, 2015 at 15:06
  • I see. The equilibrium glucose is what identifies diabetes or not. It makes sense :) But as @rumtscho mentioned, it would also be important to check the actual average levels, because if someone has a diet so horrendous that raises him to diabetic levels of sugar on daily basis, even without the disease, that is equally bad for him.
    – Luan Nico
    Commented Jul 8, 2015 at 19:19

2 Answers 2


Your blood sugar, as well as other blood markers, are in equilibrium most of the time.

A meal will create a rise in glucose, but this will only last about 2 hours, so that you will have raised glucose only 6 hours out of 24. So this is one reason why the fasting glucose is the glucose to which your body is exposed most of the time.

But more important, the doctors don't care for your peak glucose levels, they care for the equilibrium level of your glucose homeostasis, and that's 5 mmol or somewhere very close to it. If it is not there, then the mechanism for achieving it is broken, no matter what your postprandial glucose levels are. So both of your assumptions are incorrect.

And if you are on a diet which has your blood glucose levels constantly elevated, that's a very calorie dense diet and you are probably indeed in trouble. But even then, measuring your fasting glucose is important, because the doctor cannot make any conclusions from the raised levels.

By the way, there is also another test for blood sugar which gives you an estimation of not the current blood sugar, but of the average blood sugar over the last few weeks. Both an elevated fasting glucose and an elevated hgb 1ac levels are diagnostic criteria for diabetes, so if you somehow managed to keep your fasting glucose low but your average glucose unreasonably high, this could be discovered and you will be diagnosed with diabetes or warned that you are in a prediabetic condition, depending on the current diagnose guidelines used in your country. But if you suspect this, you should probably inform your physician, as I don't think hgb a1c is measured in routine blood tests.

Source: a Coursera course on diabetes I can't link because it's no longer open, sorry. Maybe somebody else has a linkable source and can edit it in.

  • That's very interesting, thanks for the reply :) So both indexes are important, but commonly the tests are to find out the equilibrium glucose levels, to which the body is exposed most of the time. If I understood correctly then, I should just need to fast for about 2 hours before the test, because that's how long the peaks take to normalize. Correct?
    – Luan Nico
    Commented Jul 8, 2015 at 19:14
  • @LuanNico the tests have been calibrated with a full night of fasting, this is why you can only rely on the result after this long fasting. First, the peak should take 2 hours to subside in healthy subjects, it may have a different speed if you have some condition. Second, this is only about first order effects, biology is usually more complicated than that. Third, these tests are measuring more than just blood sugar. I don't know how food affects each of the markers measured in a blood test, but if you are told to fast overnight, there is a reason for it, secondguessing can throw off results.
    – rumtscho
    Commented Jul 8, 2015 at 20:38

The main reason is because your body conditions change rapidly after a meal. These changes are also meal dependent (I work in a lab and we recently found a way to estimate a person's citrus fruit consumption fairly reliably based on changes in urine).

By instructing you to fast, healthcare providers are attempting to check you when your wildly fluctuating lab results are as stable as possible by minimizing factors like the timing and composition of your meal. That way any results they draw are more likely to reflect something actually wrong with your heath rather than what you last ate.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.