Why Is My Body Fat Percentage Going Up on a GLP-1?

It’s a share, not an amount

Body fat percentage is fat mass divided by total mass. That means there are two entirely different ways for it to rise: you gained fat, or you lost something that wasn’t fat. During weight loss the second is far more common, and it’s worth seeing the arithmetic once.

Say you start at 100 kg with 35 kg of fat — 35.0%. You lose 3 kg, and of that, half a kilo is fat and two and a half kilos are fat-free mass. You now weigh 97 kg and carry 34.5 kg of fat. Your body fat percentage is 35.6%. You lost half a kilo of fat, and the percentage went up six tenths of a point. Nothing went wrong; the denominator moved faster than the numerator.

The part that makes this less alarming than it sounds is what fat-free mass actually is. Fat-free mass is roughly 74% water, 19% protein and 6.5% mineral, and researchers estimate only around 60% of the fat-free mass people lose during weight loss comes from skeletal muscle. So a drop in the non-fat part of you is mostly water, and a percentage that rose because of it is not evidence that you’ve lost a lot of muscle.

The number you’re reading has an error bar

Before you interpret a change, check that the change is bigger than the noise. Measured against a reference DEXA scan, a research-grade impedance device was off in body-fat percentage by about four points on average and over-predicted fat-free mass by roughly 2.8 kg — while producing near-identical readings test to test.

That combination is the whole story of a bathroom scale: consistently wrong rather than randomly wrong. It’s why the absolute number is close to meaningless and the direction of travel, under identical conditions, still carries some information. It’s also why a move from 31.2% to 32.4% is not a finding. It’s inside the margin of a device that was never able to tell you the true figure in the first place.

Consumer scales are generally no better than the research-grade device that produced those numbers, so if anything, treat your own margin as wider.

What actually moves the number week to week

Impedance estimates how easily a small current passes through you, and water conducts. So anything that changes your fluid balance changes the reading — how much you drank, what you ate, the salt in it, where you are in your cycle, whether you trained yesterday, whether you weighed before or after the toilet.

On a GLP-1 there’s an extra reason to distrust individual readings: you’re often drinking less, and the digestive side can mean losing fluids. To be clear, nobody appears to have studied whether that specifically skews scale readings in people on a GLP-1 — treat it as a mechanical reason for caution rather than an established finding. The scale-accuracy guide goes through the fluid problem properly.

The practical consequence is that a single week’s move is mostly a report on your hydration, not your body.

What to do instead of reacting to it

Change nothing on one reading. Standardise the conditions — same time of day, first thing, after the toilet, before food or drink — then look at a monthly average rather than individual numbers, and judge the line over three to six months.

Then stop asking one device to answer the question alone. Put a waist measurement beside it, which isn’t affected by conductivity at all, and put your strength numbers beside that. If your percentage drifted up a point this month while your waist held and you added reps or load, the percentage is the least trustworthy of the three and it’s the one to discount.

The thing most people want from this number — am I keeping the muscle? — is not a question a bathroom scale can answer at the precision people try to read it at. What you can track honestly is whether the work is getting done and whether you’re getting stronger.

When it’s a question for your healthcare provider

A percentage on a bathroom scale is not a health assessment, and it isn’t the basis for changing anything medical. Take it to your healthcare provider if you’re losing weight faster than you or they expected, if you’re noticeably weaker in ordinary tasks like stairs or carrying shopping, or if something else has changed alongside the number that’s worrying you.

Anything about your dose, your schedule, or whether to keep going belongs in that conversation too — not in a reading from a device that carries a four-point margin.