Ask people what a GLP-1 does to cholesterol and most will say it falls, dramatically, the way weight does. The trial tables printed in the labels say something different and more interesting. Triglycerides fell substantially. HDL rose modestly. And total cholesterol and LDL, the two numbers people are usually told to worry about, moved by only a few percent, in some dose groups by an amount the trial could not distinguish from placebo at all. That gap between expectation and data is what this page reports.
The caveat printed in the label itself
Start with the footnote, because the label's own table tells you how to read it. The lipid parameters in the Wegovy table were not included in the pre-specified hierarchical testing. In plain terms: the trial was designed to answer other questions, chiefly weight, and these lipid measurements were collected and reported alongside the results rather than tested as endpoints in the formal sequence. They are descriptive numbers, honest and carefully gathered, but not formally tested results.
That matters for how much weight the numbers can carry. A descriptive table tells you what happened to these measures in these participants. It does not carry the statistical standing of a pre-specified endpoint, and the label is careful to say so. Everything below inherits that caveat, and it is a mark of the document's honesty rather than a reason to dismiss the figures.
Triglycerides moved the most, by a wide margin
In the Wegovy weight-reduction table, participants began with a triglyceride average of 139.5 mg/dL. On placebo, triglycerides fell 3.2 percent from baseline. On Wegovy, they fell 18.3 percent, a relative difference from placebo of 15.6 percent. That is the largest lipid movement in the table by a distance, and the pattern repeats in the tirzepatide data.
Zepbound's tables report differences from placebo with 95 percent confidence intervals, and a separate table shows the full lipid panel: triglycerides at a difference of -25.1 percent (interval -30.9 to -18.9), non-HDL cholesterol at -12.6 percent (-15.9 to -9.3), LDL at -9.6 percent (-13.7 to -5.4), total cholesterol at -6.4 percent (-9.0 to -3.6), and HDL rising by 9.3 percent (4.5 to 14.2). Even in that table, where every interval sits clear of zero, triglycerides moved roughly twice as much as total cholesterol. The ordering is consistent everywhere: triglycerides first, non-HDL and LDL in the middle, total cholesterol last, HDL up.
Total cholesterol and LDL: a few percent, sometimes less
Now the numbers people expect to move. In the Wegovy table, total cholesterol fell 4.6 percent on the medicine against 1.9 percent on placebo, a difference of 2.8 percent, from a baseline average of 156 mg/dL. LDL fell 5.3 percent against 3.1 percent on placebo, a difference of 2.2 percent, from a baseline of 78.5. HDL rose 4.9 percent against 0.6 percent on placebo, a difference of 4.2 percent. These are small shifts, reported honestly, and they are not the story most readers have heard.
The Zepbound dose-group tables add a wrinkle that deserves a careful reader. Across the three dose groups, the differences from placebo for total cholesterol were -2.1 percent (-4.5 to 0.4), -2.7 percent (-5.1 to -0.2) and -4.6 percent (-6.8 to -2.2). For LDL they were -2.9 percent (-6.6 to 0.9), -4.0 percent (-7.5 to -0.5) and -5.5 percent (-8.9 to -2.0). A confidence interval is the range of differences the data is compatible with, and when that range crosses zero, the trial cannot rule out that there was no difference at all. The first of the three total cholesterol intervals crosses zero, and so does the first LDL interval. That does not mean there was no effect. It means the trial could not establish one at those doses, and the label prints exactly that. The term is defined in GLP-1 Words Decoded: A Plain-English Glossary.
GLP 1 Tracker AppDoses, weight and how you felt, dated and private on your phone. When the lab recheck comes around, the record of what changed and when is what makes the appointment specific.
Get the app
What these numbers are not
Three boundary lines keep this page honest. First, these are percent changes from particular baselines in particular trial populations, reported descriptively. They are not a prediction for any reader, and no individual's panel will be consulted against them here. Second, they are not the cardiovascular outcome indications. Those approvals came from trials that counted heart events over years, a different kind of evidence entirely, covered at GLP-1s and the Heart: What the Outcome Trials Showed. A lipid change is a measurement; an outcome trial result is a count of things that happened to people. The two are frequently conflated, and they should not be.
Third, the modest size of these shifts is itself worth noticing against the surrounding noise. Blood pressure tells a similar story in the same trials: modest group-level falls, reported at GLP-1s and Blood Pressure: What the Trials Measured, alongside a small rise in resting heart rate, covered at Why Your Resting Heart Rate May Be Higher. The lipid picture is one panel in that same gallery: real, measurable, modest, and routinely retold as dramatic.
The conversation that follows the recheck
There is a practical consequence of all this, and it sits firmly with prescribers. People who take lipid-lowering medicines and then lose substantial weight may have their panels rechecked, because weight loss changes the picture those medicines were prescribed into. What any rechecked value means, and whether anything changes as a result, is a decision for the clinician who ordered the test, made with the full history. Blood Work on a GLP-1: Numbers Worth Keeping covers keeping that record straight, and it declines, as this page does, to interpret anyone's results.
The one-sentence version of this whole page: in the trials' own tables, the big lipid movers were triglycerides and HDL, and the celebrated pair, total cholesterol and LDL, moved modestly at best, with some dose groups indistinguishable from placebo. Anyone whose mental model says otherwise is not wrong to be surprised. The tables are right there in the label, and they are worth reading before the headlines.
Frequently asked questions
Do GLP-1 medicines lower cholesterol?
Not in the dramatic way most people expect. In the Wegovy and Zepbound trial tables, total cholesterol and LDL moved by only a few percent relative to placebo, and several of the Zepbound confidence intervals for those two cross zero, meaning the trial could not rule out no difference at all. Triglycerides moved the most by a wide margin, and HDL rose modestly. These are group-level descriptive changes, not a prediction for any reader.
How much do triglycerides drop on Wegovy or Zepbound?
In the Wegovy weight-reduction table, triglycerides fell 18.3 percent from baseline on the medicine against 3.2 percent on placebo, a relative difference of 15.6 percent. In a separate Zepbound table, the difference from placebo for triglycerides was 25.1 percent. Triglycerides moved more than any other lipid in both trials, with the usual caveat that these are group averages in particular trial populations rather than forecasts for individuals.
What does it mean when a confidence interval crosses zero?
A confidence interval is the range of values a trial's data is compatible with. When the interval for a difference between medicine and placebo crosses zero, the data cannot rule out that there was no difference at all. Two of the three Zepbound dose-group intervals for total cholesterol do exactly that, as does the lowest dose group's LDL interval. It does not mean there was no effect; it means the trial could not establish one at that level of precision.
Are the lipid changes from GLP-1 trials the same as the heart benefits?
No, and confusing the two is the most common mistake in this area. The cardiovascular outcome indications came from trials that counted actual events over years. The lipid tables are percent changes from baseline measured alongside those programs and reported descriptively. A lipid change is a measurement, not an outcome. What the outcome trials showed is a separate subject covered elsewhere on this site.
I take a statin and I am losing weight on a GLP-1. Should my dose change?
That is a prescriber question, and this page will not gesture at an answer. What can be said generally is that people on lipid-lowering medicines who then lose substantial weight may have their panels rechecked, and what any number means belongs to the clinician who ordered it. Never change, pause or stop a lipid-lowering medicine on the strength of weight loss or of anything read here.