Weight trend and scale dynamics

Reading the signal beneath the noise. Your morning scale number is mostly water and food, with fat contributing very little day to day, which is exactly why Strathlon draws a trend instead of asking you to react to a single reading. This is the deep-dive behind the Weight trend chart in your Trends (the solid actual line versus the dashed predicted line) and the weekly slope tile in your Review. In the app, tap the (i) on those sections to land here.

Signal vs noise: the one idea the whole thing hangs on

The scale measures total mass: fat, muscle, bone, organs, gut contents, stored carbohydrate and all your body water, weighed together in one number. It cannot tell those apart. That single limitation is the root of nearly every scale frustration.

Here's the asymmetry that matters. Fat mass changes only through sustained energy imbalance (calories in versus calories out, accumulated over many days), and it moves slowly. Losing even half a kilogram of actual fat represents thousands of calories of deficit banked over a week or more. But your total body water can shift by a comparable or larger amount within a single day: from food still moving through your digestive tract, from stored carbohydrate and the water bound to it, from sodium, from hydration, from hormones, and more.

So on any given morning the number on the scale is mostly the sum of these fast-moving fluid and gut-content shifts, with the slow fat-change signal hidden underneath. The fat-loss signal is real but small. The water-and-food noise is large but temporary, and it largely cancels itself out over time. Averaging across days is what lets the persistent direction (your actual fat trend) become visible. That is the entire reason a rolling trend line exists.

How big is the daily swing, really

Everything above says the daily number is noisy. The obvious next question is how noisy, and it deserves a real answer rather than a shrug, because getting the size wrong in either direction leads you astray. Told the swing is tiny, you panic over a normal Tuesday. Told it is enormous, you wave away a genuine kilogram of regain as "just water".

The best figure comes from the modelling work that underpins estimating energy intake from body weight, which measured and validated day-to-day body-weight fluctuation in free-living people as normally distributed with a standard deviation of about 0.5 kg.^[5] A measurement-theory analysis approaches it from the other side and puts the reference change value for body mass at 2.5% at 95% probability, which for a 75 kg adult is about 1.9 kg.^[6] Those two agree with each other. Put together they give the shape you should carry around:

• An ordinary day sits within roughly half a kilogram either side of your trend. That is the swing you should expect to see most mornings and should never react to.
• One to two kilograms is the outer envelope, the occasional day rather than the typical one. It happens after a big salty meal, a heavy carbohydrate day, a long flight or a bad night, and it clears.
• Anything larger, sustained over a week or more, is signal. A trend that has moved 2 kg and stayed there is not noise, and treating it as noise is the mistake in the other direction.

And the composition of those swings has actually been measured. When the actual make-up of two-week free-living weight change was analysed, 84% of a unit change in body weight was fat-free mass, and the energy density of changes of one to three kilograms averaged about 2,380 kcal per kilogram, against roughly 7,700 kcal per kilogram for pure fat.^[7] That is direct evidence for the thing this whole page is built on. Short-term movement on the scale is overwhelmingly not fat.

It is also why the averaging matters rather than merely helping. The same modelling work found that estimating a change in energy intake from body weight to a useful precision needs daily weights over more than 28 days.^[5] The signal is genuinely buried, and a run of readings is what digs it out.

Why daily weight jumps around: the five mechanisms

The scatter in your daily readings isn't measurement error and it isn't your body misbehaving. It's normal physiology. Five sources do most of the work.

1. Food volume and gut contents

Everything you eat and drink has mass, and it registers on the scale until it's absorbed, metabolised or eliminated. A large meal, a salty dinner, or a high-fibre day can leave measurable mass in transit for a day or more. This is entirely separate from any fat or energy consequence of the food. A big dinner does not mean you "gained fat overnight." That's physically impossible in energy terms over a few hours. The added weight is the meal's mass plus its associated water, and it clears. The fix isn't to fight this variable but to remove it: weigh under consistent conditions, the classic being first thing in the morning, after using the bathroom, before eating or drinking.

2. Glycogen and its bound water

Glycogen is your body's stored form of carbohydrate, held in muscle and the liver, and it is stored together with water. Each gram of glycogen is bound to roughly three to four grams of^[1][8] water, a well-established ratio. Eat more carbohydrate and glycogen stores refill, pulling water in with them, and the scale rises. Eat low-carb, train hard, or start a diet and glycogen depletes, releasing that water: the scale drops.

This is the single biggest reason for the fast "whoosh" of early weight loss when someone starts a diet, especially a lower-carb one. Much of week-one loss is glycogen water. Actual fat contributes only a small share of it. It's also why weight can jump up after a high-carb day or a refeed with no fat gain whatsoever. Rapid multi-day swings tied to changes in carb intake are glycogen and water, and they stabilise within a couple of days of returning to normal eating.

3. Sodium and fluid balance

Sodium is the main electrolyte governing how much water your body holds in the space around your cells, and the body tightly regulates its concentration. Eat more sodium (a salty meal, restaurant food, processed food) and the body transiently retains more water to keep that concentration in range, so the scale rises. Over the next day or two the kidneys excrete the excess sodium and the water leaves with it. This is normal homeostasis. It isn't a malfunction. A salty dinner the night before a weigh-in can add visible water that has nothing to do with fat, and it self-corrects. There's no need to aggressively cut sodium to "beat" the scale. Just recognise the pattern and let the trend average it out.

4. Sleep and stress

Short sleep and psychological or physical stress are associated with transient changes in fluid balance and appetite regulation. Stated carefully: sleep loss and stress can be associated with elevated cortisol. Cortisol has some mineralocorticoid-like activity, and it can be associated with transient fluid retention. Poor sleep is also associated with changes in hunger and satiety signalling that can make sticking to a plan harder. Note the hedging in that list: these are tendencies. None of them is a switch that "erases your results."

So a stressful week or a run of bad nights can show up as a temporary scale bump (mostly water) and, indirectly, as harder appetite control, both of which resolve. That's a masking effect on the scale. It isn't evidence fat loss stopped. If the scale ticks up during a high-stress or low-sleep stretch, treat it as likely-temporary water, keep logging, and let the trend line speak once things normalise. Your sleep chart is there precisely so you can check this factor before assuming your plan is broken.

5. The menstrual cycle

For people who menstruate, hormonal fluctuations across the month are a major, cyclical source of water-weight noise. Because it's such a common source of confusion, it has its own section below.

How to read the trend line: the practical skill

The core instruction is simple: read the rolling trend over one to two weeks. The daily points are supposed to bounce around: that scatter is the water, food and hormone noise described above. The solid actual line is what smooths through them to show your real direction.

Why does a rolling average work? Because the daily noise sources are largely random in direction and self-cancelling over time. A salty Tuesday pushes the number up. A glycogen-depleted Thursday pushes it down. Average across enough days and these cancel, exposing the slow, persistent fat-change signal underneath. That is exactly what a moving-average trend line is for.

How many days is "enough"? Because the noise can be as large as a week or more of genuine fat change, you generally need one to two weeks of trend before a direction is trustworthy. A two- or three-day move tells you almost nothing on its own. This is also why weighing more often can actually help rather than hurt: more data points feed the average, so the trend settles faster, as long as you're reading the line and not each morning's dot. In a large observational analysis of frequent self-weighers, daily weighing tracked weight loss, and breaks longer than a month tended to coincide with gain.^[9] And because weigh-ins from a scale that syncs to Apple Health now arrive in Strathlon automatically, that trend densifies on its own: every reading feeds the average without you logging a thing.

The dashed predicted line: what divergence means

Alongside your solid actual line, Strathlon draws a dashed predicted line. This is the weight path your logged energy deficit should produce over time, based on the well-established principle that sustained energy imbalance drives fat change^[3]. In plain terms: "if the numbers you entered are complete, and your energy expenditure is close to what we estimate, here's roughly where your weight should head."

This is a model. It's not a promise. The prediction rests on two inputs that are each estimates: the calories you log in, and the calories you expend, which any app or device can only estimate, never truly measure. Real energy expenditure varies between people and can drift over time. The dashed line is a reasonable expectation. It's not a guarantee. When your actual line drifts off it, that's a diagnostic prompt. It isn't a verdict.

When actual drifts below predicted: in rough order of likelihood

1. A logging gap (by far the most common). Un-logged bites, drinks, cooking oils and dressings, weekend meals, or simply underestimated portions. This is the first thing to check, and it isn't a personal failing: portion estimation is genuinely hard for everyone. See energy balance for how the deficit itself is built.
2. Water retention masking the trend. Sodium, glycogen refill, hormones, stress or sleep, or a hard new training block can hold water and flatten the actual line temporarily even while fat is being lost underneath.
3. A shifted maintenance level (the least common). Estimated calorie needs can be off, or true expenditure can drift (moving less on rest days, or over long diets, some modest metabolic^[4] adaptation). That shifts where "maintenance" actually sits, so the same logged deficit produces less loss than predicted. Strathlon already re-centres your calorie and protein targets on your ~7-day weight trend, so the part of this that's simply from carrying less weight is absorbed automatically as the trend falls. A stubborn, well-logged gap beyond that is the cue to revisit the target itself.

Every one of these is ordinary and fixable. None means "the method is broken" or "my body is broken." The single most likely cause is simply that the logged input isn't capturing everything eaten. So check logging accuracy first, then consider water and timing, then reconsider the calorie target, in that order.

How water masks fat loss: how to read a plateau

Fat loss and water retention are independent processes happening at the same time. That's why the scale can stay flat, or even rise, for days to a couple of weeks while your body fat is genuinely decreasing. If you lose a little fat over a week but simultaneously retain a similar amount of water (from a new training stimulus, higher sodium, hormones, glycogen, or stress), the net scale number doesn't move, even though your body composition improved. The fat is gone. Water has temporarily taken its place on the scale.

A well-known example: starting or intensifying resistance training can cause temporary water retention in muscle, associated with tissue repair and glycogen changes. A beginner can lose fat yet see the scale hold steady for a couple of weeks because of this. It is normal and temporary. The way to see through it is to corroborate with non-scale signals: your strength and performance trending up in Strathlon, how your clothes fit, tape measurements, progress photos.

Reading a plateau honestly

Given how large water noise is relative to a week's genuine fat change, it's entirely expected that the trend line will sometimes go flat for a week or two even while fat loss continues underneath: the water offset simply hasn't cleared yet. A stretch of one to three weeks of roughly flat weight is normal variance. It isn't a stall. These commonly resolve on their own with patience. The classic pattern is a flat run followed by a sudden drop as retained water releases: the "whoosh" again.

When might it be a genuine stall? Stated honestly and without alarm: if the trend is truly flat for several weeks and your logging has been accurate and complete, then a real plateau is plausible, and the levers are the usual ones: re-check portions and logging, reassess your activity, or revisit the calorie target because maintenance may have shifted. Persistent, unexplained plateaus despite accurate adherence are worth a chat with a professional. But the default correct action for a short flat stretch is patience. A drastic cut isn't warranted.

Body recomposition: a flat scale while the body changes

"Recomp" is losing fat and gaining or maintaining muscle at the same time, so your body composition improves while the scale barely moves. Fat mass and muscle mass are regulated by different processes, so in a modest deficit with adequate protein and resistance training, it's possible to lose fat while building or preserving muscle. If the fat lost and the muscle gained are similar in mass, the scale stays flat even though you're visibly leaner and stronger.

An honest scope note: simultaneous fat loss and muscle gain is most pronounced in specific groups, namely beginners, people returning after a layoff, and those with higher starting body fat. For a well-trained, already-lean person, doing both at once is slower and harder. Worth stating so nobody feels the process has failed if their scale keeps dropping as expected instead.

This is exactly why weight should never be read in isolation. A flat weight trend plus rising strength in the app is a positive signal: it's the fingerprint of recomposition. Muscle and fat differ in density and metabolic role. A leaner, stronger body at the same weight is a real, meaningful change the scale simply can't see. Pair your weight trend with strength progress to read the true story.

The menstrual cycle and water weight

For people who menstruate, cyclical hormonal changes cause a predictable, recurring pattern of water-weight fluctuation across the month, most noticeably a rise in the days leading up to a period (the late luteal phase), which then resolves once the period begins. This is one of the clearest real-world illustrations of why single-day readings mislead.

The mechanism: carefully, because online sources get this wrong

The fluid retention itself is well-observed and real. The precise hormonal mechanism, however, is more complex than the popular "progesterone triggers aldosterone, which makes you retain water" story, which is an oversimplification, and partly backwards. In fact, progesterone is a mineralocorticoid-receptor antagonist^[2]: it opposes aldosterone's water-retaining action. The net premenstrual retention appears to arise from a tangled interplay (including compensatory rises in aldosterone, progesterone metabolites that can act as mineralocorticoid agonists, and estrogen's own effects on the body's fluid-regulating systems) rather than one clean single-hormone pathway. In short, the net effect is established, though the exact mechanism is complex, so don't trust the tidy one-line version you'll see repeated everywhere.

How much, and for how long

Be careful with numbers here. The commonly cited average premenstrual water-weight increase is modest: for many people, somewhere in the region of half a kilogram to around one-and-a-half kilograms. Some individuals experience as much as two to three kilograms, but that's an upper-end, individual-variation figure: read it as "some people see as much as", well short of "normal to expect." Plenty of people notice little or none. A heavier, bloated feeling is commonly reported alongside the scale change. This is water weight, and it's temporary: no fat is involved, and it typically resolves within days once menstruation starts.

So for menstruating users, a scale rise in the premenstrual window is expected and not a setback. Where possible, compare the same phase across cycles (this month's late-luteal week against last month's) rather than comparing across phases within a single month. This connects directly to Strathlon's cycle-aware, "adapts not tracks" approach: the physiology is exactly why the weight signal has to be read across the cycle, and never against a single point in it. See cycle phase for how the app factors this in.

Myths worth correcting: a quick reference

• "The scale tells me my fat change day to day." No. Daily change is mostly water and food.
• "A big meal or cheat day means fat gained overnight." Physically impossible in energy terms over hours. It's mass, water and glycogen refill.
• "Fast week-one loss is fast fat loss." Largely glycogen water, especially on lower-carb diets.
• "Salt makes you fat / water weight must be eliminated." Water shifts are normal, temporary homeostasis.
• "Poor sleep or stress will erase my results." Overstated: they can nudge water and adherence. They don't convert a deficit into fat gain.
• "Actual below predicted means nothing works / my metabolism is broken." Usually a logging gap or temporary water. Metabolism is the least likely cause.
• "A flat scale means no fat lost." Water can mask real fat loss for a week or two or more. Check strength and measurements.
• "One flat week is a plateau, so slash calories." One to three weeks flat is normal variance. Patience first.
• "The scale not dropping means I'm failing." Flat weight plus rising strength means fat replaced by muscle. That's success.
• "Premenstrual weight gain means my diet failed." Predictable, temporary fluid retention that resolves in days.
• "Progesterone triggers aldosterone to retain water." Oversimplified and partly backwards: the net effect is real, the mechanism is complex.

Using it in Strathlon

Pulling it together, here's how to read your weight sensibly:

• Read the solid line and the weekly slope. The solid actual trend line and the weekly-slope tile are the truth of your weight direction. The daily dots are expected to jump around: that scatter is physiology, nothing more.
• Give a direction one to two weeks before trusting it. Because the noise can be as large as a week of genuine fat change, short moves mean little. Let the average settle.
• Weigh under consistent conditions. First thing, after the bathroom, before eating or drinking, removes the food-and-hydration variable rather than fighting it.
• Treat divergence from the dashed line as a prompt. It isn't a verdict. Check logging accuracy first, then water and timing, then the calorie target, in that order.
• Don't overreact to a flat stretch. One to three weeks flat is normal. The "whoosh" often follows. Patience beats a panicked cut.
• Read weight alongside its neighbours. Pair the weight trend with strength progress, energy balance and, where relevant, cycle phase. A flat scale with rising strength is a win. It isn't a stall.

The single most useful habit is the simplest one: watch the trend, ignore the noise. That's the whole reason the solid line and the weekly-slope tile earn their place: they turn a jumpy, misleading daily number into an honest picture of where you're actually headed.

References

Every specific figure, threshold and named mechanism above is sourced below. Where a claim reflects public guidance rather than a single trial, the citation is to the body that issued it. Where the evidence is genuinely mixed, the entry says so.

1. Olsson KE, Saltin B. Variation in total body water with muscle glycogen changes in man. Acta Physiologica Scandinavica. 1970;80(1):11–18. The measurement behind the "roughly 3 grams of water per gram of glycogen" figure, and the reason a low-carb week produces a fast scale drop that is mostly water. PubMed 5475323
2. Stachenfeld NS. Sex hormone effects on body fluid regulation. Exercise and Sport Sciences Reviews. 2008;36(3):152–159. Reviews how oestrogen and progesterone act on the systems that hold and release fluid: the basis for saying the net premenstrual retention is real while the single-hormone story told online is not. PubMed 18580296
3. Hall KD. What is the required energy deficit per unit weight loss? International Journal of Obesity. 2008;32(3):573–576. Examines exactly when the 3,500 kcal-per-pound rule of thumb is appropriate. The model predicts "a larger cumulative energy deficit is required per unit weight loss for people with greater initial body fat", so applying one flat figure to everyone overstates the loss for some people and understates it for others. PubMed 17848938
4. Rosenbaum M, Leibel RL. Adaptive thermogenesis in humans. International Journal of Obesity. 2010;34(Suppl 1):S47–S55. The reference work on the body defending its energy stores during sustained restriction, a partial offset that slows loss rather than the "starvation mode" that halts it. PubMed 20935667
5. Hall KD, Chow CC. Estimating changes in free-living energy intake and its confidence interval. The American Journal of Clinical Nutrition. 2011;94(1):66–74. The source for day-to-day body-weight fluctuation being normally distributed with a standard deviation of about 0.5 kg, and for the finding that daily weights over more than 28 days are needed to estimate a change in energy intake to a useful precision. PubMed 21562087 · PMC3127505 full text
6. Cheuvront SN, Ely BR, Kenefick RW, Sawka MN. Biological variation and diagnostic accuracy of dehydration assessment markers. The American Journal of Clinical Nutrition. 2010;92(3):565–573. The measurement-theory route to the same answer: a reference change value for body mass of 2.5% at 95% probability, which is the basis for treating one to two kilograms as the outer envelope rather than the ordinary day. PubMed 20631205
7. Bhutani S, Kahn E, Tasali E, Schoeller DA. Composition of two-week change in body weight under unrestricted free-living conditions. Physiological Reports. 2017;5(13):e13336. The direct measurement of what short-term weight change is actually made of: 84% fat-free mass per unit change, and an energy density of about 2,380 kcal per kilogram against roughly 7,700 for pure fat. PubMed 28676555 · PMC5506524 full text
8. Kreitzman SN, Coxon AY, Szaz KF. Glycogen storage: illusions of easy weight loss, excessive weight regain, and distortions in estimates of body composition. The American Journal of Clinical Nutrition. 1992;56(1 Suppl):292S–293S. The source for glycogen being stored in hydrated form with three to four parts water, which is the mechanism behind the fast "whoosh" of early diet weight loss. PubMed 1615908
9. Helander EE, Vuorinen AL, Wansink B, Korhonen IK. Are breaks in daily self-weighing associated with weight gain? PLOS ONE. 2014;9(11):e113164. The observational evidence that frequent self-weighing tracks weight loss, and that breaks of longer than a month tend to coincide with gain. Support for weighing often and reading the line rather than the dot. PubMed 25397613 · PMC4232563 full text

This is general educational information rather than medical advice. The physiology described here is well-established in broad terms, but individual variation is large, and mechanisms are framed as tendencies and associations rather than certainties. Body composition, fluid balance and menstrual patterns can be affected by medications and by conditions such as thyroid, kidney, heart, PCOS and others. Anyone with unexplained rapid weight change, persistent swelling, a persistent unresponsive plateau despite accurate logging, or a diagnosed condition should consult a qualified clinician. See our Terms for more.

All stats · User guide · Support · Home