Fasting Before a Blood Test: Complete Guide

Which tests require fasting?

Can I drink water while fasting?

What you CANNOT have during fasting

• Tea or coffee, even without sugar. Caffeine and milk affect test results
• Fruit juice: high sugar content spikes blood glucose
• Any food: biscuits, fruits, anything solid
• Chewing gum: stimulates insulin release
• Alcohol: affects liver enzymes and triglycerides for 24–48 hours

What about medications?

Best time to get a fasting blood test

What happens if I accidentally ate?

Tell the lab technician and your doctor immediately. For fasting blood sugar and lipid profile, eating beforehand will give falsely elevated results. Most labs will reschedule rather than draw an inaccurate result. For tests that don't require fasting, eating is not a problem at all.

Special situations

• Diabetics: Ask your doctor specifically whether to take insulin or diabetes tablets before a fasting test
• Children under 5: Prolonged fasting can cause hypoglycaemia, ask your paediatrician for specific guidance
• Pregnant women: Keep the fasting period shorter (6–8 hours maximum) and eat immediately after
• Thyroid medication (levothyroxine): TSH is ideally measured before taking the morning dose

References

Sources cited on this page. PubMed links open the original abstract.

1. Lippi G, Salvagno GL, Montagnana M, Lima-Oliveira G, Guidi GC. Quality standards for sample collection in coagulation testing. Semin Thromb Hemost. 2012;38(6):565–575. PMID 22669845 · doi:10.1055/s-0032-1315961

Why fasting matters for certain blood tests

Eating a meal causes significant short-term changes in blood chemistry that would render certain test results uninterpretable. Fasting for 8–12 hours allows the body to return to a stable metabolic baseline:

• Triglycerides: The test most sensitive to recent food intake. After a fatty meal, the intestine releases chylomicrons (triglyceride-rich lipoprotein particles) into the bloodstream, elevating triglyceride levels dramatically – sometimes 5–10× higher than fasting values. A fasting lipid panel is required for accurate triglyceride measurement and for LDL calculation (which uses the Friedewald equation, requiring fasting triglycerides below 4.5 mmol/L for validity).
• Glucose: Fasting glucose above 7.0 mmol/L on two occasions meets WHO criteria for diabetes. Non-fasting glucose above 11.1 mmol/L with symptoms also diagnoses diabetes, but in the borderline range (5.6–7.0 mmol/L) a fasting sample is required. An oral glucose tolerance test (OGTT) requires a 10-hour fast and precise glucose loading.
• Insulin and C-peptide: These must be measured fasting – insulin rises dramatically postprandially, making fasting values the meaningful baseline for insulin resistance assessment and C-peptide testing in suspected insulinoma.
• Iron and transferrin saturation: Iron levels rise significantly after eating red meat. Fasting is required for iron studies – particularly important when diagnosing haemochromatosis, where a fasting transferrin saturation above 45% is the screening threshold.

Tests that do NOT require fasting

Many commonly ordered tests are minimally affected by recent food intake and can be measured at any time of day – saving patients the inconvenience of fasting:

• HbA1c: Reflects average glucose over the prior 2–3 months (the lifespan of red blood cells). Completely unaffected by recent meals – it is the preferred diabetes screening test precisely because fasting is not required.
• Full blood count (FBC): Blood cell counts are not meaningfully affected by eating.
• Thyroid function (TSH, T4, T3): Thyroid hormones are stable throughout the day regardless of food. Note: TSH has a diurnal rhythm (slightly higher at night and in the early morning) but this variation is small and clinically insignificant for standard interpretation.
• CRP and ESR: Inflammatory markers are not food-dependent.
• Kidney function (creatinine, eGFR, U&E): Minor post-meal increases in creatinine occur with high-protein meals (meat) – for precision, avoid large meat meals the day before. Routine clinical eGFR does not require fasting.
• Total cholesterol and HDL: Can be measured non-fasting; only triglycerides and LDL (calculated) require fasting. Some guidelines now permit non-fasting lipid panels for cardiovascular risk assessment, using non-HDL cholesterol (total minus HDL) as the risk marker when triglycerides are not needed.