Do You Need to Fast Before a Blood Test? What a Study Found

A study in the November 2026 issue of JAMA Internal Medicine looked at millions of routine lab results and asked how much the hours since a patient's last meal really change them. Only a handful of tests shifted meaningfully when people had fasted for less than 8 hours, mainly glucose, triglycerides, gamma-glutamyl transferase and lipase. Between an 8 to 12 hour fast and a longer one, no test showed a meaningful difference.

This is about fasting before a blood draw, which is a short pause from food so that a meal does not distort a lab reading, and not about fasting as a health practice. The practical advice does not change: follow the instructions of whoever ordered your test.

What the study did

Researchers in Korea went back through the laboratory records of one hospital and compared results by how long each patient had gone without eating.

Type of study	Retrospective, single-center, cross-sectional
Where	One 670-bed secondary-care university hospital in the Seoul metropolitan area, Republic of Korea
When	Samples collected between February 2021 and December 2023
Who	Adult outpatients (18 or older) with a documented fasting duration
Size	9,755,547 test results from 101,148 patients across 372 test items
Fasting groups	Less than 8 hours, 8 to 12 hours, more than 12 hours

Fasting duration was worked out from the time between the patient's last meal and the blood collection.

The team did two kinds of comparison. First, they compared the three fasting groups with each other, adjusting for age, sex and hospital department. Second, among patients who had been tested more than once, they compared each person with themselves at different fasting durations. The second approach matters because people who turn up fasted and people who do not may differ in other ways.

To decide whether a difference was worth caring about, they used established analytical performance specifications: published limits, based on natural biological variation, for how much error a lab result can carry before it stops being acceptable.

What it found

Few tests were affected. The group comparison flagged 27 analytes (the substances being measured) with differences beyond those limits. Only 10 of them held up when patients were compared with themselves. The authors take that gap as a sign that many of the apparent differences came from confounding, not from fasting.

Four stood out. With less than 8 hours of fasting, these were higher:

• Glucose, by 10.8%
• Triglycerides, by 20.1%
• Gamma-glutamyl transferase (a liver enzyme), by 14.8%
• Lipase (a pancreatic enzyme), by 36.4%

Longer than 12 hours added nothing. No analyte exceeded the limits between the 8 to 12 hour group and the more than 12 hour group.

Within the first 8 hours, timing mattered for some tests. Looking at the under 8 hour group in 30-minute steps, glucose and lactate decreased significantly as fasting time increased. Lipid concentrations showed no significant linear trend.

The authors conclude that the findings "support analyte-specific rather than uniform fasting requirements". In plain terms: fasting matters for some tests and not for most, so a blanket rule may ask more of patients than the results need.

What it does not show

• It is observational. Nobody was assigned to fast or eat. The study read existing records, and the authors themselves found that confounding explained many of the first-pass differences.
• It comes from one hospital. One center in Korea, with its own patients, meals and lab methods. Other settings may differ.
• It is about groups, not about you. An average shift across thousands of patients does not tell you what your own result would do after a particular meal.
• It does not record what people ate. The abstract describes time since the last meal, not the meal.
• It does not change your lab's rules. The study argues for more targeted requirements. It is evidence for the people who write those rules, not permission to ignore the ones you have been given.

What this means before your next test

Follow the instructions from the doctor or lab that ordered the test. They know which tests are on the form, and this study shows precisely why that matters: glucose and triglycerides are sensitive to a recent meal, and those are among the most commonly ordered tests. If you were told to fast and you ate, say so when you arrive instead of hoping it will not show. If you are unsure whether fasting is needed, or what you may drink, ask beforehand.

Do not read this study as saying you can skip fasting. It says a few results shift when the fast is shorter than 8 hours, and that is exactly the situation a fasting instruction is there to prevent.

If you fast regularly

If you practice intermittent or extended fasting, the opposite mistake is the one to watch for. A long fast before a test is not what the lab asks for. In this study, going beyond 12 hours made no meaningful difference to any result compared with 8 to 12 hours, so there is nothing to gain from turning a blood draw into a longer fast.

Tell your doctor about your fasting routine. Reference ranges and fasting instructions assume an ordinary overnight gap since your last meal, and this study did not examine the multi-day fasts described on pages like the 72-hour fast. Your doctor can only interpret your numbers properly if they know how long you had really gone without food. For a general picture of what changes as the hours pass, see the fasting stages guide and the fasting schedules overview.

Fasting is not right for everyone, and this post is general information, not medical advice. If you are pregnant or breastfeeding, have diabetes or take blood-sugar medication, have a history of eating disorders, take other regular medication, or are under 18, talk to a doctor before fasting. The same people should check before changing how they prepare for a test, because fasting can interact with medication timing.

Knowing your hours with QoFast

QoFast is a fasting tracker app for iPhone and Android, for intermittent fasting and multi-day extended fasts. Its timer works offline and survives restarts, and the fasting history records when each fast started and ended, which makes it easy to tell a doctor how long you had fasted. QoFast gives general information, not medical advice, and is free to download with an optional Pro subscription on the App Store and Google Play.

FAQ

How long should you fast before a blood test? For as long as the person who ordered the test tells you. In this study, results did not differ meaningfully between 8 to 12 hours and more than 12 hours.

Which blood tests are affected by eating? In this study, mainly glucose, triglycerides, gamma-glutamyl transferase and lipase were higher with less than 8 hours of fasting. Ten analytes in total were confirmed out of 372 test items examined.

Is a longer fast better before blood work? The study found no meaningful difference for any test beyond 12 hours compared with 8 to 12 hours.

Can I skip fasting if my test is not one of those four? That is a decision for your doctor or lab, not for you or this post. Follow the instructions you were given.

Sources

• Shin S, Kim S, Yu S, Yoo SJ, Park K, Chang J, Um TH, Cho CR, Whang DH. "Fasting Duration and Differences Across Routine Laboratory Tests." JAMA Internal Medicine, November 2026 issue. DOI: 10.1001/jamainternmed.2026.4718. PubMed: 42804179.