How do I tell a plateau from a new set point, and how long should I wait before changing anything?

Nine weeks flat at 71.6 kg, from a start of 94 kg. Semaglutide 2.4 mg, month fourteen. Intake weighed, not estimated, at 1,650 kcal. Steps averaging 9,400. Three lifting sessions weekly. Protein 125 g.

My intake is genuinely measured and my activity is genuinely tracked, so I do not think this is a logging story. What I want to understand is conceptual: is there a difference between "a plateau I could break with a further intervention" and "a new set point that my body will defend"? People use both phrases and I cannot tell whether they mean the same thing or whether one is real and the other is folklore.

Practically: how long is it reasonable to sit at a flat weight before deciding it is not going to move on its own? Nine weeks feels like a long time but I have also been told repeatedly that I change things too fast, which is fair.

I am not in distress about this. BMI is now in the normal range. I am asking because I would rather understand what is happening than keep guessing.

The two phrases do mean different things, and the distinction is real but weaker than "set point" language implies. What you almost certainly have is neither: you have an accurately measured energy balance at a weight your body is currently defending mildly, and nine weeks of flat weight at 1,650 kcal is a result, not a problem.

Plateau versus set point

A plateau is a state of energy balance. Intake equals expenditure. It is broken by changing one of the two. It carries no implication about defence and no implication about permanence.

A set point, in the strong version of the theory, is a defended value that the system actively regulates back toward through both expenditure and appetite. The strong version has poor support in humans: if it were true, populations could not have gained weight steadily for decades, and individuals could not maintain large losses, which some clearly do.

The version with better support is usually called a settling point: body weight settles where the whole system of appetite, food environment, activity and physiology reaches equilibrium, the defence is asymmetric, being much stronger against loss than against gain, and the equilibrium moves when the inputs move. Under that model your 71.6 kg is not a wall, it is where your current inputs put you, and one of your current inputs is a drug.

What is genuinely defended, and this is the part with real evidence, is the appetite and expenditure counter-regulation described in weight-loss maintenance research: elevated ghrelin, reduced leptin, and expenditure below prediction, all of which persist for a long time after loss. Those defences do not stop further loss, they raise its cost. The drug is currently paying part of that cost for you, which is precisely why the withdrawal trials show what they show.

How long to wait

Given genuinely measured inputs, the thresholds I would use:

You are at nine weeks with clean data, so you are in the "decide what you want" band rather than the "diagnose a problem" band.

What your numbers tell you

Stable weight at 1650 kcal means TDEE = ~1650 kcal/day
Predicted TDEE for 71.6 kg with your activity: roughly 1900-2100
Shortfall vs prediction: ~250-450 kcal/day

That shortfall is the combination of the drug's effects, adaptive thermogenesis after a 22.4 kg loss, and the reality that population prediction equations are poor for individuals. It is worth knowing, because it means a formula-based maintenance target would over-feed you by 300 kcal and you would drift upward while believing you were maintaining. Your measured number beats any equation.

What I would actually do at a normal BMI

Deliberately spend three to six months at maintenance before deciding anything else. Reasons:

• Maintenance is the skill that determines the outcome, and it is a different skill from losing. Practising it while you still have pharmacological support is easier than practising it later without.
• Counter-regulatory signalling attenuates with time at a stable weight. Sitting at a weight for six months makes it cheaper to hold than it is at week nine.
• A maintenance phase is when you can actually raise protein and total energy enough to train properly, which is when body composition improves even at a flat weight. Recomposition at maintenance is more available than at a deficit.
• If you later decide you want another few kilograms, entering that from a well-fed, well-trained baseline works better than extending a fourteen-month deficit.

If instead you want to keep losing, change one input, wait four weeks, reassess. My order of preference: add steps, then add a training session, then reduce intake by 150 kcal. Intake last, because at 1,650 kcal and 125 g of protein you do not have much room before energy availability becomes the problem you are solving instead.

Worth raising with a clinician at a normal BMI: what the plan is for how long you continue, and at what dose. That is a different conversation from the one you were having at 94 kg and it is the right time to have it.

On the water-masking question in a longer-horizon case like this, since nine weeks is long enough that most of the usual fluid explanations have expired: at month fourteen the fluid candidate that survives is not glycogen or training, it is chronically elevated cortisol from a very long deficit.

The pattern to look for, and it is recognisable once you know it: weight flat or creeping for weeks, then a sudden drop of 1.5-2.5 kg over three or four days, usually after a rest week, a holiday, a few nights of good sleep, or a deliberate few days at higher intake. That is not fat loss appearing suddenly; it is retained fluid releasing when the stress input drops. It means the fat loss had been happening all along and was being masked.

If you suspect that, the diagnostic is a deliberate diet break: seven to ten days at your measured maintenance plus 200-300 kcal, keep protein and training the same, keep weighing. Two possible outcomes and both are informative:

• Weight rises 0.5-1 kg then falls below the previous plateau by day ten. Fluid masking confirmed, the deficit was working, and you can resume with more confidence and probably less restriction than you thought you needed.
• Weight rises and stays up. You were at genuine balance and the plateau is real.

Either way, ten days costs you very little at your rate of loss, and it buys a clean answer. The reason people resist it is psychological rather than physiological: eating more on purpose while stuck feels like giving up. It is the opposite; it is the only way to get an uncontaminated measurement.

One measurement caveat specific to being at a lower weight, which nobody has mentioned and which affects how you read your own trend line.

At 94 kg, a 0.5 kg weekly loss was 0.53% of your body weight per week and stood out clearly against roughly 1-1.5 kg of daily fluctuation. At 71.6 kg, the same 0.5 kg would be 0.70% per week, but the loss you can now sustainably achieve is smaller in absolute terms, perhaps 0.2-0.3 kg per week, while your daily fluctuation has barely shrunk.

Signal at 94 kg:    0.5 kg/week against ~1.2 kg noise -> visible in ~3 weeks
Signal at 71.6 kg:  0.25 kg/week against ~1.0 kg noise -> visible in ~4-6 weeks

So the minimum observation window to detect real change gets longer as you get lighter, purely as a statistical matter. Nine weeks flat at 71.6 kg is roughly as informative as five weeks flat at 94 kg would have been. That is an argument for the patience thresholds set out above rather than against them, but it is also a reason not to feel that nine weeks is an unreasonably long time to have waited. It is about the right length of time.

The corollary is that as you approach a goal, you should be extending your evaluation windows rather than shortening them, which is the opposite of what impatience suggests.