Reading Your Own Labs

Patients arrive at follow-up appointments with a lab report they have already read, usually on a phone, usually at eleven at night, and usually having Googled one value out of context and frightened themselves.

I would rather you read it properly. So here is what I actually look at, when, and what each thing is telling you.

This is not a substitute for having a clinician interpret your results against your history. It is so that you can arrive at that conversation able to participate in it.

The schedule

Blood pressure and waist circumference belong on this table too, at every point. They are free, they move early, and they are more informative than most of the blood work.

The markers that move first

If the scale has stalled at month three and you want evidence that something is working, look here.

Triglycerides

Often the earliest meaningful change, and frequently dramatic. Triglycerides track closely with visceral fat and insulin sensitivity, and both improve early. Falls of thirty to fifty percent by three months are common.

If your triglycerides were elevated and are now normal, something substantial has changed in your metabolism regardless of what the scale says that week.

ALT and liver enzymes

Mildly elevated ALT is extremely common and almost always silent. It is the usual laboratory footprint of metabolic dysfunction-associated steatotic liver disease — fat in the liver, driven by visceral adiposity and insulin resistance.

It also improves early, because hepatic fat is among the first fat to go. Semaglutide has been studied specifically in steatohepatitis with encouraging histological findings, and in ordinary practice a normalising ALT is one of the most satisfying things to watch.

If you have ever been told incidentally that you have a fatty liver and filed it away as unimportant, this is the marker to watch.

The ones that take longer

HbA1c

HbA1c reflects roughly the preceding three months of glucose exposure. This is the single most common misunderstanding I encounter: someone checks it at six weeks, sees little change, and concludes the drug is not working.

Check it at three months at the earliest, and read it at six. A move from 5.9% to 5.4% — the top of prediabetes to comfortably normal — is a meaningful change in trajectory, and it is the kind of thing that alters what the next twenty years look like.

HDL cholesterol

Rises, but slowly, over six to twelve months rather than three.

LDL cholesterol

Frequently barely moves, and this surprises people who expect every lipid to improve together. LDL is much more strongly determined by genetics and by saturated fat intake than by weight. If your LDL was the problem, weight loss alone may not fix it, and that is not a failure of the drug.

The ones that go the wrong way

Two deficiencies actually show up in practice, and both are consequences of the arithmetic: food volume falls by half, requirements do not fall at all.

Ferritin

Iron stores. Depleted before haemoglobin drops, which is why a normal CBC does not reassure you and why you have to ask for ferritin specifically.

More likely if you menstruate, if you are vegetarian or vegan, or if you have had bariatric surgery. Worth finding, because it is trivially correctable and because it is a plausible contributor to fatigue and to the hair shedding that turns up around month four.

B12

Same logic, slower onset. Worth checking annually, and sooner if there is any neurological symptom.

What to bring to the appointment

The single most useful thing you can do is bring the trend rather than the value.

A ferritin of 28 means very little on its own. A ferritin that has gone 61 → 42 → 28 across a year means something specific and it means it immediately. One number is nearly uninterpretable; four points is a conversation.

I keep my own baseline, three-month, six-month and twelve-month results alongside the doses and measurements in the Zenday App, which is mostly about not having to reconstruct any of it from four different patient portals the night before an appointment. Whatever you use — a spreadsheet is genuinely fine — the point is that the trend exists somewhere you can see it in one view.

I have sat on the clinician side of this for twenty-one years and I will tell you plainly: a patient who can show me their trajectory gets a better appointment than a patient who tells me they think it has been going well. Not because I doubt them. Because we can make an actual decision instead of a guess.

Three things worth saying about ranges

A “normal” result is not a good result. Reference ranges describe the middle 95% of a reference population, which in the case of things like fasting insulin is a population with a great deal of metabolic disease in it. Being inside the range is not the same as being where you want to be.

Two of these tell you about your liver, and there is a third worth calculating from them. Ask for a FIB-4 score — it uses age, ALT, AST and platelet count, all of which are already on this panel, and it is the number that separates common harmless fatty liver from the version that scars. GLP-1s and fatty liver disease explains what to do with the result.

Creatine changes one of these. If you take creatine, serum creatinine rises by roughly 0.1–0.3 mg/dL without any kidney injury, which drags calculated eGFR down with it — say so before anyone interprets the panel, and see GLP-1s and kidney disease for why that matters here.

A single abnormal value is usually not an emergency. Labs vary between draws, between labs, with hydration, and with what you ate. A repeat is often the right next step rather than a diagnosis.

Two more worth asking for by name. Lipoprotein(a) is genetically determined, does not respond to weight loss, and only needs measuring once in your life — cholesterol and lipids on a GLP-1. And if you take levothyroxine, a TSH about 6–8 weeks after any substantial weight change, because the dose was calculated for a heavier body — thyroid disease on a GLP-1.

Numbers are not the same as outcomes. A moved lab value is a surrogate, and surrogates do not always translate into the thing you actually care about — how to read a GLP-1 study covers the distinction.

Your direction matters more than your position. Someone moving from 6.1% to 5.6% is in a much better situation than someone sitting flat at 5.5% and rising, even though the second number looks better on the day.

That is the whole argument for measuring on a schedule rather than measuring when you are worried. The trend is the information. The single value is just the most recent point on it.

Two results on this list have a specific confounder worth declaring before anyone interprets them: creatine raises creatinine (GLP-1s and kidney disease), and high-dose biotin interferes with thyroid and troponin immunoassays (the best supplements for GLP-1 hair loss).