How to test, and how to tell if it changed
On this page
You usually don’t need to fast. The 2026 US guideline accepts a test after eating, and a normal meal moves LDL only about 8 mg/dL. To tell whether a change was real, test at the same lab, the same way, 4 to 12 weeks apart. Two LDL results can differ about 19% with nothing changed.
What we’d do
Pick one lab and stick with it. Test the same way each time, fasting or not, and note which. After a food change, wait 6 to 12 weeks for LDL and 4 for triglycerides. Then compare the two results against the noise. An LDL drop under about 19% may be chance, and a drop of 29% or more almost certainly isn’t. If a home test surprises you, check it at a lab before you act on it.
If you remember three things
- Fasting is optional for most people. A meal raises triglycerides up to about 26 mg/dL and lowers LDL about 8.
- One result isn’t the truth. Your body’s own day-to-day swing plus the lab’s error can move LDL by a fifth.
- Same lab, same way, 4 to 12 weeks. That’s how you see a real change.
Why one cholesterol test fools people
A result comes back as a crisp number, so it feels exact. People retest after a month of oatmeal, see their LDL drop from 142 to 128, and decide it worked. Or they see it rise and panic. Both reactions assume the first number was the truth.
Your cholesterol moves on its own from week to week, and every lab measurement carries a small error. Put the two together and two tests of the same person, with nothing changed, can land a long way apart.
Like stepping on a scale
You wouldn’t judge a diet from one morning on the scale. Water, salt and the time of day move it a pound or two, so you look at the trend. A cholesterol panel works the same way, with one difference. You get one number every few weeks instead of one every morning. That makes it more important to know how far a single result can wander.
How far a result can move with nothing changed
In 1993 four researchers pooled 30 studies, published from 1970 to 1992, of people who had their cholesterol tested again and again. Each person’s own levels swung on average by 6.1% for total cholesterol, 7.4% for HDL, 9.5% for LDL and 22.6% for triglycerides. Add the most error the CDC allows a certified lab, and you get the size a change has to reach before you can trust it.
| Marker | “Probably real” (80% sure) | “Real” (95% sure) | At a typical level |
|---|---|---|---|
| Total cholesterol | 12% | 19% | At 220: 27 and 41 mg/dL |
| HDL | 15% | 23% | At 50: 8 and 12 mg/dL |
| LDL | 19% | 29% | At 130: 24 and 37 mg/dL |
| Triglycerides | 42% | 64% | At 150: 63 and 96 mg/dL |
An LDL of 130 can come back at 106 with nothing changed. That’s why a single food change of 7 to 13 points often hides inside the noise on its own, and why two or three together give you a result you can see. A good lab usually does better than the CDC’s limit, so these bands are on the wide side.
Put your own two panels in below and it stamps each change as “Can’t tell yet”, “Probably real” or “Real change”.
Do you need to fast?
Usually not. The guideline recommends “a standard nonfasting or fasting lipid profile”. In 33,391 adults in Copenhagen, a normal meal changed the numbers by at most:
- Triglycerides up about 26 mg/dL (0.3 mmol/L), for up to six hours
- Total cholesterol and LDL down about 8 (0.2 mmol/L), for three to five hours
- HDL down about 4 (0.1 mmol/L)
- ApoB, non-HDL and the total to HDL ratio not at all
The guideline suggests fasting if a parent or sibling had early heart disease or a cholesterol disorder, if you have a known triglyceride problem, or if your triglycerides came back at 400 or more after eating. At 400 the 1972 LDL formula can’t be used, so a fasting retest gives a better LDL. Whichever you choose, test the same way each time, so a meal doesn’t fake a change.
When to test
| Who | How often |
|---|---|
| Adults | From age 19, at least every 5 years, more often with risk factors |
| Children | Once at 9 to 11 |
| Relatives of someone with FH or early heart disease | From age 2 |
| After a food change | 6 to 12 weeks for LDL, 4 weeks for triglycerides |
| After starting or changing a medicine | 4 to 12 weeks later, then every 6 to 12 months (your prescriber sets this) |
Things that move your result for a while
- Fast weight loss. The guideline lists rapid weight loss among the causes of a higher LDL. Retest once your weight has held for a few weeks.
- Pregnancy. Cholesterol rises in pregnancy on purpose, so the usual lines don’t apply. Take your numbers to your midwife or doctor.
- Menopause. LDL jumps in the year either side of the last period (why cholesterol goes up after menopause).
- A serious infection. The guideline suggests repeating triglycerides measured during a severe infection.
- A different lab or equation. Certified labs can differ by a few percent on their own, and a lab that switched from the 1972 LDL formula to a newer one can shift your LDL with nothing changed in you.
Are home tests as good as a lab?
Some home meters are on the CDC’s list of certified systems for total cholesterol and HDL, and the FDA’s consumer page (last updated in 2018) says a home test can be about as accurate as your doctor’s if you follow the directions carefully. But the certification tests are run by trained staff on a lab bench, and a finger you prick yourself at home is a different situation. And the home meters that show an LDL calculate it from the other numbers, the same way most labs do. Use a home result to keep an eye on things, and confirm any surprise at a lab. Are home cholesterol tests accurate? goes through the evidence.
How to get the cleanest comparison
- Write down how you tested. Fasting or not, the lab, the date, and anything unusual that week.
- Make one change at a time if you can. Then you know what worked.
- Retest the same way. Same lab, same fasting, similar time of day.
- Compare against the band. If the change is inside it, keep going and test again rather than giving up.
- Look at non-HDL too. It’s measured numbers only, so it doesn’t carry the LDL formula’s error.
7 sources
- Blumenthal RS, Morris PB, Gaudino M and others, 2026 ACC/AHA guideline on the management of dyslipidemia, JACC and Circulation, 2026doi.org
- Langsted A, Freiberg JJ, Nordestgaard BG, Fasting and nonfasting lipid levels: influence of normal food intake on lipids, lipoproteins, apolipoproteins, and cardiovascular risk prediction, Circulation, 2008pubmed.ncbi.nlm.nih.gov
- Smith SJ, Cooper GR, Myers GL, Sampson EJ, Biological variability in concentrations of serum lipids: sources of variation among results from published studies and composite predicted values, Clinical Chemistry, 1993pubmed.ncbi.nlm.nih.gov
- CDC Clinical Standardization Programs, Cholesterol Reference Method Laboratory Network: analytical performance criteria for tests used to assess cardiovascular disease riskcdc.gov
- CDC Clinical Standardization Programs, List of CDC-certified cardiovascular disease assayscdc.gov
- US Food and Drug Administration, Cholesterol: home use testsfda.gov
- Martin SS and others, Comparison of a novel method vs the Friedewald equation for estimating low-density lipoprotein cholesterol levels from the standard lipid profile, JAMA, 2013pubmed.ncbi.nlm.nih.gov
