Reading the trial chart for your dose

Under the trial chart on the Progress page, Tirzing reads the chart back to you in words: with your dose taken into account. This page explains what those sentences mean and, just as importantly, what they don’t.

There is no “cohort” for your dose

A trial arm is named for the dose it ends up on, not the dose it’s on at any given moment. Everyone in SURMOUNT-1’s 10 mg arm started at 2.5 mg and climbed through 5 mg and 7.5 mg to get there. So if you’re on 7.5 mg, there is no “7.5 mg cohort” you can be winning or losing against: the study never randomised one. “Which cohort am I?” is the wrong question, and most confusion about this chart starts there.

Instead we line you up against each arm three different ways, and every sentence says which one it’s using. Mixing them up is the easiest way to talk yourself into a result you haven’t got.

The three comparisons

• Same week: where each cohort was at the same number of weeks since starting. This is what the chart itself draws.
• Same rung of the ladder: where a cohort was after the same number of weeks on the dose you’re on now. If no arm ever used your dose, we say so rather than invent a comparison.
• Same total drug: you’ve taken a certain number of milligrams so far; we find the week each cohort had taken about the same amount, and tell you where they were then.

These can disagree, and that’s not a bug: they answer different questions. You can be level with one cohort on the calendar while sitting ahead of another one’s time at your dose. Neither is “the” answer.

Headroom

We also tell you how far you are from the top of a cohort’s dose ladder, and how long they’d already been at their maintenance dose by your week. This is the sturdiest thing on the panel: it’s plain arithmetic rather than an inference. Sitting mid-ladder means you still have steps in reserve; a cohort that topped out at week 12 doesn’t.

Why there’s no “bang for your buck” number

It’s tempting to divide weight lost by milligrams taken. We deliberately don’t show that, because the answer is rigged: response to these drugs flattens out as the dose climbs, so the lowest arm always wins on that measure. It would tell every member, on every dose, that they’d get better value from less medication, which is a property of the curve’s shape, not a finding about you, and not a thing a tracking app should be nudging. The “same total drug” comparison captures the same idea honestly.

What these sentences are not

• Not a verdict. Trial curves are cohort averages. Roughly half of any cohort sits above its own average, so edging past one isn’t beating anybody: individual results vary far more than the gaps we show.
• Not precise. Differences under 2 percentage points are reported as “tracking with”, because the published curves simply aren’t sharp enough to resolve them. The trial figures are read off charts, the arms we plot include people who stopped taking the drug, and a single weigh-in can swing a couple of pounds on water alone.
• Not a prediction. An early position tells you little about a final one. In one SURMOUNT-1 follow-up analysis, 90% of the people who had lost under 5% by week 12 still reached 5% or more by week 72: it just took them longer.
• Not causal. Nothing here can tell you that a dosing choice caused your result. You’re one person without a control group; you might simply be a strong responder.

As with everything else here: this is for your own tracking, not medical advice. Dose changes are a conversation for you and your provider.