I keep seeing people throw around STEP trial numbers without context, so I wanted to do a proper deep dive for anyone interested in the actual science behind semaglutide's efficacy data.

The STEP (Semaglutide Treatment Effect in People with obesity) program consists of multiple trials. Here's a breakdown of the key ones:

Some critical points that get overlooked:

1. STEP 2 shows lower weight loss in T2D patients. If you have diabetes, expecting 15% body weight loss may be unrealistic — 10% is more typical. This is likely because insulin resistance itself promotes fat storage.
2. STEP 3 combined sema with intensive behavioral therapy and got 16% — suggesting that lifestyle changes ADD to the medication effect. This is not "just take a shot and eat pizza."
3. STEP 4 is the scary one. Patients who discontinued semaglutide after 20 weeks regained about two-thirds of the weight they'd lost within a year. This has massive implications for the "is this a lifetime medication?" question.
4. STEP 5 shows durability. For those who stayed on it, the weight loss was maintained at 2 years. The curve doesn't keep going down forever — it plateaus around month 12-15 — but it doesn't bounce back either.

Happy to discuss any of this in more detail. I have the full papers if anyone wants specific data points.[1-5]

This is excellent. The STEP 4 data is what people need to understand before starting. When I started semaglutide, I went in with the assumption this would be a long-term medication, possibly lifelong. And I'm at peace with that.

The framing of "you have to take it forever" as a negative always bothers me. I also take lisinopril for blood pressure "forever." Nobody clutches their pearls about that.

Can you talk more about the responder analysis? I've heard that the "average" weight loss is misleading because the distribution is bimodal — some people lose a ton, some barely respond.

Previously posted:

Can you talk more about the responder analysis? I've heard that the "average" weight loss is misleading because the distribution is bimodal — some people lose a ton, some barely respond.

Great question. The distribution isn't quite bimodal but it is wide. Here's the responder breakdown from STEP 1:

So about 1 in 3 people lost >20% of their body weight — that's a super-responder group. But also ~14% didn't even hit 5%, meaning they were essentially non-responders.

The factors that predict response aren't fully understood, but early weight loss (first 12-16 weeks) is a strong predictor of long-term outcomes. If you're not losing at least 5% by week 16 on the full dose, the likelihood of achieving significant long-term weight loss is lower.[1]

Fantastic thread. One important nuance about the STEP trials that I always bring up with patients:

The trials used an intention-to-treat (ITT) analysis, which includes patients who dropped out. In the "treatment policy" estimand, dropouts are included at their last observed weight. In the "trial product" estimand (patients who actually stayed on the medication), the results are even better — about 17-18% weight loss for completers in STEP 1.

This matters because when YOU take semaglutide and stick with it, your expected outcome is closer to the completer analysis than the ITT analysis. The ITT is dragged down by people who discontinued early.

Also worth noting: the placebo group lost 2.4% body weight because ALL participants received lifestyle counseling. The drug effect is the difference: about 12.5 percentage points attributable to semaglutide itself.