Weight loss plateau or noise: how to tell them apart

A plateau is a change in rate

The scale can repeat the same number for ten days while the trend underneath keeps sliding. It can also bounce around enough to hide a trend that really has stopped. So the useful question is about the slope: has the trend’s weekly change moved toward zero, and has it stayed there?

Slowing and stopping are different things. A trend that eases from 1 lb/wk to 0.3 lb/wk is still moving, and that kind of taper is normal in a campaign that runs for months. A trend whose rate hovers around zero for weeks on end has leveled.

Why the range settles the question

Every estimate of the trend carries uncertainty, and the rate carries more of it than the level does. Say it reads −0.4 lb/wk ±0.5. That spans everything from a loss of almost a pound a week to a slight gain, so the readings can’t yet separate a slowdown from a stall.

Watch what happens as the readings pile up. If the range tightens and still straddles zero, the flattening holds up. If it tightens on the loss side, the trend had kept falling all along and the flat readings were scale noise. Sparse logging keeps the range wide for longer, which is why a stall that shows up during a stretch of two weigh-ins a week takes more time to confirm. The forecast range guide covers how that uncertainty carries forward into a date.

How many weeks to wait

Short stalls often come from water. Glycogen, the carbohydrate kept in your muscles and liver, is stored along with a good deal of water, so a change in how much of it you carry moves the scale for days at a time (Kreitzman, 1992).

Give it at least three weeks of regular weigh-ins before treating a flat trend as a plateau. That window spans more than one water swing, and it gives the rate enough readings for its range to narrow. If something in your routine changed during that time, write it down where you can see it next to the line, since a new job or a stretch of travel often explains a pause better than physiology does.

Why the trend bends even on a steady routine

Even when intake and activity hold perfectly steady, the trend curves. A smaller body needs less energy to run and to move around, and the gap between what you eat and what you burn shrinks with every pound. Modeling work from the NIH found that weight takes about a year to cover half of its eventual change once eating habits shift and hold there, and then settles at a new level (Hall et al., 2011).

Two adaptive changes push in the same direction:

• Energy use drops a bit more than size alone predicts. People keeping off 10% or more of their weight burned roughly 10–15% less than their new body composition would suggest (Rosenbaum and Leibel, 2010).
• Appetite rises. In a 52-week trial, people ate about 100 kcal a day above their starting intake for each kilogram they lost, an effect more than three times larger than the change in energy use (Polidori et al., 2016).

A review by NIH researchers describes the familiar result in weight-loss trials: rapid early loss that stalls after several months, with rising appetite as the main reason (Hall and Kahan, 2018). A flattening trend six months into a campaign matches the shape those trials report under a fixed routine.

What a leveling forecast means

A forecast drawn as a straight line will always reach the goal eventually, because a falling line keeps going. A curved forecast can do something a line can’t: it can settle. When the fitted curve’s resting point lands higher than your goal, the readings are saying that the current pace leads to a new steady weight short of where you’re aiming.

That’s useful to know early. It separates a plan that is on its way from one that is coasting toward a level, and it gives you a reason to revisit the goal and its pace before months pass. Whether to change the routine or the goal itself is your call, and anyone on a medical treatment should bring the question to the clinician who manages it.

Reading a flat trend in Balance Point

The Today screen shows the trend’s change per week with its range beside it, in a form like “−0.8 lb/wk ±0.4”, and a rate chart below the main one draws the same uncertainty across time. When the trend changes direction, a reversal notice appears once on Today.

Balance Point’s preferred forecast is a fitted curve whose resting weight is left open. It waits until a campaign’s weigh-ins span 16 weeks and number at least 20, and the readings also have to show a visible curve. Once that model is active, the forecast may say the trajectory settles above your goal. Earlier in a campaign it falls back to a constant rate with its interval drawn in, and it drops the date entirely when the last few weeks hold too little data. Notes sit on the chart too, so a flat stretch can be read beside the week a routine changed.