Letter says fast from 10pm, the nurse says black coffee is fine, my mother says nothing after dinner including water. Do you actually need to fast for blood tests now?

Annual bloods booked for the end of the month, the boring core panel I've settled on after my adventures in over-testing (documented elsewhere on this board). And somehow the simplest test on the menu has produced three contradictory briefings.

The booking letter: "fast from 10pm, water only". The nurse, when I rang to move the time: "oh, for what you're having black coffee is fine, don't worry too much". My mother, unprompted, when I mentioned it: nothing after dinner, INCLUDING water, that's how she's done it for forty years and her generation had proper results, apparently. A colleague completes the set by insisting nobody fasts anymore and the letter is just a template nobody's updated.

I don't especially mind skipping breakfast. What I mind is doing the wrong version and getting a flagged number that's really just toast, then spending three weeks chasing a phantom like I did with the MRI findings. So which is it in 2026? What still genuinely needs an empty stomach, is black coffee actually neutral or nurse-politeness, and is my mother's dry fast doing anything besides making her miserable?

I can report from the far end of your mother's method. January, my first full check, I fasted fifteen hours to be extra safe and skipped water too because my mother-in-law has the same rule. Nearly fainted in the chair, and it took two goes to find a vein. The nurse was very kind about it and very clear: the water rule isn't just allowed, it's encouraged, being dehydrated makes your veins harder to find and can concentrate some of the numbers they're trying to read.

The other thing she told me that nobody had mentioned: my diabetes check wasn't a fasting glucose at all, it was HbA1c, which she described as an average of the last three months. It doesn't care what you ate since Easter, let alone at 10pm. Fifteen hours of noble suffering, largely decorative.

January check, flask of coffee in the car, drunk the second the needle came out, best coffee of my life. The nurse's ruling for mine: tea with milk breaks a fast, black coffee is "a grey area", water is boring but bulletproof. I went with boring.

Fi, the direct answer: for routine screening panels, fasting has quietly become the exception rather than the rule, and every one of your four briefings contains a piece of the truth, including, in one narrow sense, your mother's.

The big change is lipids. Since a major European consensus statement in 2016, non-fasting lipid panels have been standard in much of routine practice, because total cholesterol, HDL and LDL barely move after a meal; triglycerides rise, but only by about 0.3 mmol/L on average, and the reference ranges were adjusted to match. The reason fasting was ever required is the part almost nobody explains: for decades LDL wasn't measured, it was calculated with the Friedewald equation, which uses your triglyceride value and becomes unreliable when triglycerides are high, as they are after food. Labs now measure LDL directly or use newer equations, so the overnight fast lost its original job. There's also a fair argument that a fed sample better represents the state your arteries actually live in for most of every day. What still earns a true fast: a formal fasting glucose (8 or more hours, water only), although HbA1c has taken over much of routine diabetes screening precisely because it needs no fasting, as Nathan discovered; a repeat fasting lipid panel when a non-fasting triglyceride comes back very high, above roughly 4.5 mmol/L; and a short list of specialised tests where the lab will say so explicitly. Separately, a few tests care about the clock more than your stomach, iron, cortisol and testosterone all swing through the day, which is why some draws are morning-only appointments.

On coffee: black coffee is essentially calorie-free but not biologically inert, caffeine can nudge glucose and free fatty acids, so for a strict fasting glucose the protocol is water only, and for a non-fasting panel it's a non-issue. Both the nurse and the flask man are right, for different tests, which is exactly the point: the instruction that wins is the one from your own testing service, because it's written for the specific assays they run, and if the letter and the phone call disagree, ask them to confirm which applies to your panel. That two-minute call is cheaper than the phantom-chasing you're rightly wary of. Your mother's dry fast, finally: the food half of her discipline was once orthodox, but the no-water half never helped and actively works against her, hydration makes the draw easier and mild dehydration can skew a few results. The background on how the fasting-or-not question fits into reading the panel is in the site's guide to cholesterol and lipid screening.

One addition from the running side: the fast isn't the only prep that matters. I once gave blood the morning after a hard 10k and a liver enzyme came back flagged, cue a week of quiet dread before my doctor explained that hard exercise leaks the same enzymes from muscle and asked, with a straight face, whether I'd considered resting. Rest day before blood day is a rule I keep now. The fast gets all the attention and the leg day gets none.

Done, and closing the loop. Rang the clinic as instructed: for my panel, no fast required at all, the letter is indeed a template "older than some of the staff" (nurse's words, delivered with feeling). Went hydrated, rested, caffeinated. One cheerful needle, nothing flagged, no phantoms to chase.

Reported all this back to my mother, who listened carefully and announced that forty years of doing it properly wasn't going to be undone by one phone call. Some reference ranges never shift. Thanks all, particularly for the Friedewald bit, it's pleasingly rare to learn WHY a rule existed before you're allowed to stop following it.