The Drug That Might Work - If We Slow Down Long Enough to Find Out
Is Ozempic the way to go? Maybe, maybe not, we just don't have enough data.
The most prescribed drug in America right now has less than three years of long-term outcome data.
Nobody is saying that out loud.
GLP-1 receptor agonists - semaglutide, tirzepatide, the entire class - have taken over the cultural conversation about weight loss with a speed that should make every clinician uncomfortable.
Ozempic is a household name. Wegovy is on the cover of magazines. Celebrities are not-so-quietly crediting it. The stock prices have reflected a gold rush. And somewhere in the middle of all that noise, the actual science - careful, longitudinal, outcomes-based science - is still catching up.
I am not here to tell you GLP-1s are dangerous. The data does not support that.
I’m also not here to tell you to go out and get some either. The data also does not support that.
I am here to tell you we do not know enough yet. And in medicine, that distinction matters enormously.
The Rush to Prescribe Is Outrunning the Research
Here is what we actually know.
GLP-1 receptor agonists work. In the SURMOUNT-1 trial, tirzepatide produced mean weight reduction of approximately 22.5% at 72 weeks. The STEP 1 trial for semaglutide showed roughly 14.9% body weight reduction over 68 weeks.
Those are real numbers. Clinically meaningful. For patients with obesity-related comorbidities - cardiovascular disease, type 2 diabetes, joint deterioration - this class of drug has genuine, evidence-backed utility.
But here is what the headlines are not telling you.
The longest continuous outcome data we have on semaglutide for weight management is roughly two to three years. For a drug now being prescribed to millions of people as a long-term or potentially indefinite intervention, that is not a track record. That is a promising early chapter.
We do not yet know what happens to metabolic function at year five. Year ten. We do not know what the cessation curve looks like for patients who use these drugs for a decade and then stop. We are moving at cultural speed when the science demands clinical patience.
What Rapid Weight Loss Is Actually Doing to the Body
This is where I want to slow the conversation down considerably.
When a patient loses 15 to 22 percent of their body weight in 68 to 72 weeks, they are not losing fat alone. They are losing muscle. They are losing bone density. That is not a side effect. That is physiology.
The data on lean mass loss during GLP-1 therapy is real and it is underreported in the popular press. Studies tracking body composition during semaglutide treatment have shown that a significant portion of total weight lost - estimates range from 25 to 40 percent depending on the study and protocol - is lean mass, not adipose tissue. For a 50-year-old patient who is already dealing with age-related sarcopenia, that is not a neutral trade.
I operate on joints. I see what sarcopenic bodies look like - not on a DEXA scan but on an operating table. Weak, atrophied muscle around a joint is not just an aesthetic issue. It is a functional one, a structural one, and a surgical risk factor. Patients with low lean mass have worse outcomes after joint replacement. They recover more slowly. They fall more. The downstream consequences of aggressive muscle loss are not hypothetical to me.
Bone density tells a similar story. Rapid weight loss - regardless of mechanism - is associated with reductions in bone mineral density. We know this from bariatric surgery literature going back decades. GLP-1-driven weight loss may be less aggressive than surgical intervention, but the mechanism is not entirely different. The STEP trials were not powered to detect fracture risk as a primary endpoint. That does not mean the risk is absent. It means we have not looked hard enough yet.
And the rebound data - what happens when patients stop - is genuinely concerning. The STEP 4 trial showed that patients who discontinued semaglutide regained approximately two-thirds of their lost weight within one year. That is not a treatment effect. That is dependency. And if the drug is producing muscle and bone loss on the way down, and then weight regain on the way back up, the patient may end up in a worse body composition position than where they started.
We need to say that clearly.
The Emerging Case for a Different Approach
Here is where I want to stake a directional position, with full transparency that the data supporting it is still thin.
The standard GLP-1 dosing protocol is designed for maximum weight loss velocity. That is what the phase three trials optimized for. It is what got FDA approval. It is what the prescribing guidelines reflect.
But maximum velocity is not the same as optimal outcome.
Microdosing - using GLP-1 agonists at sub-therapeutic or low-therapeutic doses, with the explicit goal of slower, more controlled weight loss over a longer timeline - is being discussed seriously in metabolic medicine circles.
The hypothesis is straightforward: slower weight loss, combined with structured resistance training and adequate protein intake, preserves lean mass while still delivering meaningful metabolic improvement. You get the appetite regulation and the glucagon suppression without the aggressive body composition changes that come with rapid loss.
Is this proven? No. The long-term microdosing data does not exist yet. That is the honest answer. But the physiological rationale is sound, and it aligns with everything we know from the bariatric and sports medicine literature about how to lose weight without gutting your musculoskeletal architecture.
The framework I find compelling is this: GLP-1 drugs as a metabolic bridge, not a metabolic destination.
Use them to get appetite and blood sugar regulation under control. Use the lower-dose window to build the behavioral infrastructure - resistance training, protein targets, sleep, stress management - that has to exist for any weight loss to be sustainable. Then taper or maintain at the lowest effective dose, with the explicit goal of not being on maximum-dose semaglutide indefinitely.
This is not how most of these prescriptions are being written right now.
The Question Nobody Is Asking Their Prescriber
The patients flooding into primary care offices asking for Ozempic are not asking about lean mass. They are not asking about bone density trajectories. They are not asking what happens to their body composition at maximum dose over two years, or what the discontinuation curve looks like, or whether there is an alternative protocol that trades some speed for better structural outcomes.
They are asking: “Does it work?”
And the answer - based on the weight loss numbers alone - is yes.
But that is the wrong question.
The right question is:
Work for what?
Over what timeline?
At what cost to the rest of the body?
As physicians, we are responsible for asking the question the patient does not know to ask. That is not paternalism. That is clinical leadership.
The enthusiasm around GLP-1s is not wrong. The underlying science is promising and the cardiovascular outcome data from SELECT and LEADER trials is genuinely compelling. But promising is not the same as proven. And in medicine, moving from promising to proven requires time, rigor, and a willingness to slow down when the culture is sprinting.
What I Am Watching For
I am not prescribing a protocol here. I am not qualified to give individualized prescribing guidance in a newsletter, and anyone who tells you they have the definitive GLP-1 microdosing answer right now is outrunning the data.
What I am watching for - and what I think every clinician managing patients on these drugs should be tracking:
Body composition, not just body weight. Scale weight tells you almost nothing useful. DEXA scan or bioelectrical impedance at baseline and every six months tells you whether you are losing fat or muscle. If it is muscle, the protocol needs to change.
Resistance training as a non-negotiable co-intervention. The research on preserving lean mass during caloric restriction is unambiguous - resistance training is the primary tool. If a patient is on a GLP-1 and not lifting, they are losing muscle. Full stop.
Long-term outcome data as it emerges. The five and ten year studies are coming. I am reading them as they publish. The story this drug class tells at year two and the story it tells at year seven may be very different.
Honest conversations about discontinuation. If the plan is indefinite use, say so. If the plan is to taper, build that timeline in from day one with a clear protocol for maintaining outcomes.
The drug might work. The early data suggests it does - in the short term, at the doses tested, for the endpoints measured.
But medicine has a long history of enthusiasm outrunning evidence. HRT. Arthroscopic knee surgery for osteoarthritis. Opioids for chronic pain. In every case, the early data looked promising. The cultural momentum was real. And the full picture took years - sometimes decades - to emerge.
GLP-1s may be different. I genuinely hope they are.
But hope is not a clinical strategy.
Slow down. Measure everything. And do not let a cultural gold rush substitute for a longitudinal outcome.
That is the incision point.
What are your thoughts?
Have you or someone you know been using Ozempic?
Have you/they had success?
Has there been side effects?
Just because something works does not mean we understand the full cost.
Great article. My BIL was on them for T2D and weight loss. He also had bone loss and lost several teeth. Was told it was a side effect :-(. He's not in great health overall anyway.