For generations, patients have arrived at clinics hungry, trusting that an empty stomach was the price of accurate bloodwork. A sweeping analysis of nearly ten million test results from Seoul now suggests that for the vast majority of what labs measure, that sacrifice was never necessary — a practice born of legitimate caution around glucose and lipids that quietly expanded, unchallenged, into a universal rule. The study does not overturn medicine overnight, but it places a quiet question at the center of a ritual millions perform each year: was this ever really about the science, or simply ab
Study challenges routine fasting requirement for most blood tests
Fasting may have been unnecessary for most of us all along
Mark
So this study looked at nearly ten million blood tests and found that fasting only mattered for ten of the 121 things they measured. That's a pretty small fraction. Why has fasting been the default for so long if it's not actually necessary?
Mimi
Because it started with a real problem. Lipid and glucose tests genuinely are affected by what you eat. Clinicians needed to control for that variability, so they said fast for eight to twelve hours. But then as lab panels got bigger and more diverse, the fasting requirement just came along for the ride. Nobody systematically checked whether it actually helped for all the other tests.
Luke
Right, but I want to be precise about what the study actually shows. It's from a single hospital in Seoul over three years. The findings are compelling, but they're not yet universal proof. Different populations might respond differently. And the authors themselves say this doesn't yet warrant changing the standard—they're calling for reconsideration, not immediate change.
Mark
Fair point. So what would actually need to happen for labs to change their policies?
Mimi
Other hospitals would need to replicate the findings. Professional organizations would need to review the evidence and update their guidelines. That takes time. But the momentum is clearly shifting. The commentary from Copenhagen essentially said the evidence is there—fasting isn't necessary for most tests.
Luke
And we should note: the study looked at whether results fell outside established acceptable variation ranges. That's a high bar. It's possible fasting causes small changes in some analytes that are still within acceptable limits. The study doesn't claim fasting has zero effect on everything except those ten—just that the effect doesn't matter clinically.
Mark
So for a patient right now, what should they do? Keep fasting?
Mimi
Yes. The researchers explicitly say don't disregard your doctor's instructions. This is evidence for future change, not a reason to ignore current guidelines. But it does suggest that within a few years, the rules might be different.
Luke
And it's worth noting the practical angle: fasting is a real burden for some people. Elderly patients, people with diabetes, people who work early shifts. If most of those people don't actually need to fast, that's a meaningful quality-of-life improvement, assuming the science holds up.
Le Pouls
- A blanket fasting rule applied to 121 different blood analytes turns out to matter for only 10 of them — a gap between clinical habit and clinical evidence that is difficult to ignore.
- For patients with diabetes, the elderly, and working people, the demand to go without food for up to twelve hours carries real physical and logistical costs that have never been formally justified for most tests.
- The fasting requirement was never designed to be universal — it originated for lipid and glucose panels specifically, then spread by institutional inertia as lab panels grew more complex.
- Researchers are now calling for an analyte-specific approach: fast before glucose and lipid tests, eat normally before everything else — a distinction simple enough to implement but significant enough to reshape routine care.
- Independent commentators describe the findings as strong support for a growing body of evidence, though changing entrenched clinical guidelines will require replication across multiple hospital systems and action from professional bodies.
For generations, patients have arrived at clinics hungry, trusting that an empty stomach was the price of accurate bloodwork. A sweeping analysis of nearly ten million test results from Seoul now suggests that for the vast majority of what labs measure, that sacrifice was never necessary — a practice born of legitimate caution around glucose and lipids that quietly expanded, unchallenged, into a universal rule. The study does not overturn medicine overnight, but it places a quiet question at the center of a ritual millions perform each year: was this ever really about the science, or simply about habit?
For decades, the instruction before a blood test has been simple and absolute: nothing to eat for eight to twelve hours. Millions of patients have complied without question. A new study of more than 9.7 million blood test results now suggests that for most of what labs actually measure, the rule was never necessary.
Led by Sunghwan Shin of Asan Medical Center in Seoul, the research examined data from over 101,000 adult outpatients tested between 2021 and 2023. Patients were grouped by fasting duration — under eight hours, eight to twelve, or more than twelve — and their results compared across 121 different blood analytes. The finding was striking in its narrowness: only ten analytes showed clinically meaningful differences based on fasting, primarily glucose, triglycerides, and a handful of enzyme markers. For the remaining 111 — liver enzymes, hormones, blood cell counts, vitamins — fasting made no measurable difference. No test benefited from fasting beyond twelve hours.
The origins of the blanket rule reveal how clinical habits calcify. The eight-to-twelve hour fast was originally designed for lipid and glucose panels, where food intake genuinely distorts results. But as laboratory panels expanded over the decades to include dozens of additional measurements, the fasting requirement expanded with them by default — never systematically tested, never formally validated for most of what it now governs.
The human cost is not trivial. For people with diabetes, the elderly, or those managing demanding schedules, prolonged fasting before a routine appointment is not merely inconvenient — it can be genuinely harmful. The practice has persisted largely through institutional momentum.
The Seoul team proposes an analyte-specific approach: require fasting only for tests that demonstrably need it, and let patients eat normally before everything else. Clinical biochemists at Copenhagen University Hospital, writing in an accompanying commentary, called the findings strong support for abandoning the universal fasting rule, while acknowledging that overturning entrenched practice takes time and broader replication. For now, most patients will still be told to arrive hungry. But the evidence beneath that instruction has quietly shifted.
For decades, the instruction has been routine: show up to your blood test hungry. Don't eat or drink anything but water for eight to twelve hours beforehand. Millions of patients have followed this rule without question, assuming it was essential to getting accurate results. A new analysis of more than 9.7 million blood tests suggests that assumption may have been wrong for most people.
The study, led by Sunghwan Shin, a laboratory medicine physician at Asan Medical Center in Seoul, examined blood test data collected between 2021 and 2023 from a 670-bed university hospital. The researchers divided their subjects—over 101,000 adult outpatients—into three groups based on how long they had fasted before their tests: less than eight hours, eight to twelve hours, or more than twelve hours. They then looked for meaningful differences in results across 121 different blood analytes, or measurable substances, comparing what they found against established standards for acceptable variation in lab testing.
The results were striking in their narrowness. Of the 121 analytes examined, fasting duration produced clinically significant differences in just ten. Those ten were mostly the usual suspects: glucose and triglycerides, along with gamma-glutamyl transferase and lipase. For the remaining 111 analytes—including liver enzymes, pancreatic enzymes, blood cell counts, hormones, and vitamins—fasting made little or no measurable difference to the accuracy of the results. The researchers found no additional benefit to fasting beyond twelve hours for any test they examined.
The irony is that this blanket fasting requirement was never meant to be blanket at all. It originated decades ago from legitimate concerns about lipid and glucose testing, where food intake genuinely does affect measurements. Clinicians began instructing patients to fast for eight to twelve hours before these specific tests to control for dietary variables. But over time, the practice metastasized. As lab panels expanded to include dozens of different measurements, the fasting requirement followed along, applied by default to nearly everything, even though no one had systematically tested whether it actually mattered for most of those tests. Different laboratories ended up with different recommendations, adding to the confusion.
The burden this imposes on patients is real. Fasting overnight before an early morning appointment means skipping breakfast. For people with diabetes, the elderly, or those with other health conditions, going without food for that long can be genuinely difficult or even risky. For working people, it means scheduling tests around meal times or enduring hunger during the workday. The practice has persisted largely through inertia—it's what labs have always done, so it's what they continue to do.
The Seoul researchers' findings suggest a different path forward. Rather than a one-size-fits-all fasting requirement, they propose moving toward what they call an "analyte-specific approach." In other words: fast if you're getting a glucose or lipid panel, because those tests genuinely need it. For everything else, eat normally. The distinction is straightforward enough that labs could implement it without much difficulty.
In a commentary accompanying the study, clinical biochemist Liv Tybjærg Nordestgaard of Copenhagen University Hospital and her colleagues called the findings "strong support for accumulating evidence that fasting blood sampling is not necessary." They stopped short of saying labs should immediately abandon fasting requirements across the board—the study is large and rigorous, but it comes from a single hospital system, and changing entrenched clinical practice takes time. But they made clear that the old blanket rule no longer makes sense.
The authors of the study themselves were cautious in their conclusions, noting that their findings alone may not yet be enough to overturn decades of standard practice. They urged a reconsideration of "rigid fasting requirements" in favor of a more thoughtful, targeted approach. What happens next depends on whether other hospitals and lab networks replicate these results and whether professional organizations decide the evidence is strong enough to update their guidelines. For now, most patients will still be told to fast. But the scientific ground beneath that instruction has shifted.
Citations marquantes
Traditional rigid fasting requirements may warrant reconsideration in favor of a more nuanced, analyte-specific approach to pretest preparation— Study authors
The study provides strong support for accumulating evidence that fasting blood sampling is not necessary— Liv Tybjærg Nordestgaard, Copenhagen University Hospital
Envie de l'histoire complète ?
Lire l'original sur ScienceAlert ↗