The calorie deficit, explained properly

That sentence is true. It is also close to useless as advice, and pretending otherwise is why so many people feel they have failed at something simple. "Calories in, calories out" describes a real law. Both terms in it are far harder to measure, and far harder to control, than the phrase suggests. This article explains the physics honestly, then explains why the arithmetic so often fails to predict what happens to a real person.

One thing first. This is general education, not personal advice. If you have a medical condition, take any regular medication, are pregnant, or are recovering from an eating disorder, work with your own doctor and dietitian rather than a web page. Rapid weight loss, extreme restriction and constant preoccupation with food are all reasons to speak to a doctor rather than push harder.

What is a calorie deficit?

A calorie deficit means your body is using more energy over some period than it takes in from food and drink, so it makes up the shortfall from its own stores. The word "calorie" in everyday use means a kilocalorie, the unit printed on food packets. A deficit is not a food, a supplement or a plan you buy. It is a state, and many eating patterns produce it: smaller portions, fewer meals, grilled instead of fried, no sugary drinks, or more protein so you feel full sooner. When people say a diet "works", they usually mean it made the deficit easier for them to hold. That is real and useful, but it is a claim about behaviour, not about a magic property of the food.

The law is real, and that is not the argument

Energy balance is thermodynamics. Energy is not created or destroyed inside a human body, so if the energy stored in your tissues falls, more energy left than arrived. No credible researcher disputes this. Anyone selling you a food that "burns fat regardless of how much you eat" is either confused or lying.

So the law is safe. Everything else is uncertain: how much energy went in, how much went out, how both numbers change once you try to change them, and how long the body takes to settle at a new weight.

Kevin Hall and colleagues modelled this in The Lancet. Their model of adult metabolism shows bodyweight responds to a change in intake slowly, with half times of about a year, and that people carrying more body fat lose more weight for the same change in intake but take longer to reach a stable weight. The body is not a spreadsheet that settles overnight. It is a slow system with feedback in it.

Where your energy actually goes

Your total daily energy expenditure, usually shortened to TDEE, is made of four parts. Most people have only heard of one of them.

Basal metabolic rate

Basal metabolic rate, or BMR, is what you spend simply existing: keeping your heart beating, your kidneys filtering, your brain running, your body warm. It is the largest single component of daily energy demand, as the Aberdeen research group led by Alexandra Johnstone noted in their study of 150 adults in the American Journal of Clinical Nutrition. Most people are surprised by this. The gym is not where the bulk of your energy goes.

The thermic effect of food

Digesting and processing food costs energy. Klaas Westerterp's review of diet-induced thermogenesis found that for a mixed diet eaten at energy balance, this comes to roughly 5 to 15 per cent of daily energy expenditure. It is higher on a high-protein diet and lower on a high-fat one. It is a real effect, but it is not large enough to rescue a diet on its own.

Exercise activity

This is deliberate exercise: the gym session, the run, the cycle ride. For most people who are not athletes it is a smaller slice than they imagine, and it does not simply add on top of everything else. A study of 332 adults across five populations, led by Herman Pontzer and published in Current Biology, measured total energy expenditure with doubly labelled water and compared it against measured physical activity. Expenditure rose with activity at low activity levels, then flattened out at higher ones. The body appears to hold total expenditure within a narrower range than a simple additive model predicts.

NEAT, the part almost nobody has heard of

Non-exercise activity thermogenesis, or NEAT, is everything else your body does. Walking to the bus. Standing while you talk. Climbing stairs. Cooking. Fidgeting. Shifting in a chair. James Levine, who named it, defines it in a review in the American Journal of Physiology as the energy spent on everything that is not sleeping, eating or sports-like exercise, and notes that environmental and biological factors together explain the substantial variance in it between people.

NEAT matters more than its obscurity suggests. In a 1999 experiment published in Science, Levine and colleagues overfed 16 non-obese volunteers by 1,000 kilocalories a day for eight weeks. Two-thirds of the rise in their total daily energy expenditure came from increased NEAT, not from exercise. Fat storage varied ten-fold between individuals, and the change in NEAT predicted who resisted gaining fat. Some people, when overfed, simply start moving more without deciding to.

The same group later tracked posture and movement every half second for ten days in 10 lean and 10 mildly obese sedentary volunteers. The obese participants sat, on average, two hours longer per day. That pattern did not change when the obese participants lost weight or when the lean participants gained it, which suggests it is at least partly biologically set rather than chosen.

Why do two people of the same size burn different amounts?

Because body size explains a lot but not everything. In the Johnstone study of 150 adults, fat-free mass accounted for 63 per cent of the between-person variation in basal metabolic rate, fat mass for about 6 per cent and age for about 2 per cent, which left roughly a quarter of the variation unexplained by any of the factors measured. Add to that the very large person-to-person differences in NEAT that Levine's work documented, and two people who look alike on paper can genuinely have different daily needs. This is why an online equation gives an estimate rather than a truth, and why comparing your results to a friend's is close to meaningless. It is also why the World Health Organization describes obesity as a chronic, relapsing disease arising from complex interactions between genetics, neurobiology, eating behaviour, access to a healthy diet, market forces and the wider environment, and says dietary and activity patterns are largely shaped by conditions that constrain personal choice.

"Calories in" is much harder to know than it looks

Here is where the arithmetic starts to come apart.

What people report is not what they ate

The classic study on this was published in the New England Journal of Medicine in 1992 by Steven Lichtman and colleagues at St Luke's-Roosevelt Hospital Center. They studied obese patients who repeatedly failed to lose weight despite reporting that they ate under 1,200 kilocalories a day. Ten people with a history of this "diet resistance" had their actual energy intake and expenditure measured over 14 days.

Their metabolisms were normal. Total energy expenditure and resting metabolic rate sat within 5 per cent of what their body composition predicted, and the thermic effects of food and exercise matched the comparison group. The reporting was what differed. On average they underreported food intake by 47 per cent and overreported physical activity by 51 per cent. This is a measurement problem, not a character flaw. Estimating your own intake is a hard task that almost everybody is bad at, trained professionals included.

If you are eating carefully and the scale is not moving, this is usually the first place to look. Our piece on why weight loss stalls goes through the other common reasons.

What the label says is not exactly what is in the packet

Food labels are estimates with legal tolerances built in. Under India's Food Safety and Standards (Labelling and Display) Regulations, the compliance rule published by FSSAI allows a tolerance of a maximum of minus 10 per cent of the declared value for a nutrient, at any point within the declared shelf life. That is a sensible rule for a manufacturer. It also means the number on the packet is a target, not a measurement of the packet in your hand.

Atwater factors are averages, not measurements

The energy figures on labels are not obtained by burning your particular food. They are calculated. The same FSSAI compendium lists the conversion factors used in India: 4 kilocalories per gram for carbohydrate and protein, 9 for fat, 2 for dietary fibre, and so on. These are the Atwater factors, and they are averages that work reasonably well across a mixed diet.

They work less well for particular foods. Researchers at the United States Department of Agriculture measured the energy people actually extracted from almonds in a controlled 18-day feeding study, published in the American Journal of Clinical Nutrition. The measured value was 4.6 kilocalories per gram. The Atwater calculation gave 6.0 to 6.1, an overestimate of 32 per cent. Some of the energy in a whole nut is simply not available to the person eating it.

Cooking and processing change what you absorb

How food is prepared changes how much energy you get from it. Rachel Carmody and colleagues at Harvard published a study in PNAS using mice as a model. Cooking substantially increased the energy gained from meat, producing weight changes not explained by how much the animals ate or moved, and cooking beat pounding for both meat and starchy tubers. The authors state the implication for labels directly: current food labelling systematically overestimates the caloric potential of poorly processed foods. This was animal work, so treat the size of the effect in humans as unsettled. The direction is not controversial. Softer, finer, more cooked, more processed food gives up its energy more easily.

Is a calorie a calorie?

As a unit of energy, yes, and that is the sense in which the phrase is defensible: 100 kilocalories of sugar and 100 kilocalories of paneer hold the same energy, and if all else were held equal they would affect your energy balance the same way. But all else is never held equal, because food does things besides deliver energy. It affects how full you feel, how long you stay full, how much you eat at the next meal, how much of the energy you actually absorb, how much muscle you keep while losing fat, and how likely you are to still be doing this in three months. So the honest answer is that a calorie is a calorie for the purpose of the energy books, and nothing like a calorie for the purpose of a human being trying to hold a deficit for months. Both halves of that sentence matter, and most arguments on the internet consist of one side shouting the first half and the other side shouting the second.

"Calories out" is harder still

At least the food is in front of you. Your expenditure is not.

Consumer fitness trackers give confident-looking numbers. A Stanford team tested seven wrist-worn devices against indirect calorimetry in 60 volunteers of varied age, size, skin tone and fitness, and published the results in the Journal of Personalized Medicine. Most devices measured heart rate reasonably well. None achieved an error in energy expenditure below 20 per cent. The authors advised caution about using device energy estimates in health programmes. Treat the calories-burned figure on your watch as a rough motivational number, not as an input to a calculation.

The body also adjusts as you lose weight, and it does not stop when you do. Michael Rosenbaum and colleagues at Columbia showed that the drop in energy expenditure after losing 10 per cent or more of body weight persists in people who have kept it off for over a year, and is not fully explained by their smaller body. A National Institutes of Health team followed up 14 of the 16 contestants from a televised extreme weight-loss competition six years afterwards. They had lost an average of 58.3 kilograms during the 30-week contest and regained 41.0 kilograms. Six years on, resting metabolic rate was still an average of 704 kilocalories a day below baseline, and the part not explained by body size or age came to about 499 kilocalories a day. Those who kept the most weight off carried the most metabolic slowing.

None of that makes weight loss impossible. It does mean the deficit you start with is not the deficit you end with, and that a plan built on a fixed number will drift out of date.

Does the 3,500 calorie rule actually work?

The rule that 3,500 kilocalories equals one pound of body weight is one of the most repeated numbers in dieting, and it is a poor guide to what happens over time. Kevin Hall examined it directly in the International Journal of Obesity, and found that the energy deficit required per unit of weight loss depends on how much body fat a person started with: the rule roughly matches the prediction for people with high initial body fat, but overestimates the deficit needed by people with less body fat, and the composition of the weight lost shifts as loss continues. The bigger problem is that the rule is static. It assumes your expenditure stays where it was, so it predicts that a constant reduction in intake produces steady loss forever, which is not what happens. As you get lighter you spend less, the gap narrows, and the same eating pattern produces less and less further loss until weight settles at a new level. Hall's later Lancet modelling put the half time of that adjustment at around a year, which is why a plan that looked exactly on track for six weeks can look badly off by month six without anything having gone wrong.

The professor who lost weight on convenience store food

In 2010, Mark Haub, a professor of human nutrition at Kansas State University, ran an experiment on himself. For two months he took about two-thirds of his energy from vending machine and convenience store food: snack cakes, sugary cereals, crisps, biscuits. The rest came from a daily protein shake, plus some vegetables and a multivitamin. He held himself to roughly 1,800 kilocalories a day, well below what he had been eating. He lost about 27 pounds, his body mass index moved from the overweight range into the normal range, and, to general surprise, his cholesterol and triglyceride readings improved. The story was reported by MinnPost and by NPR.

Be precise about what this shows. It shows that a calorie deficit drives weight loss. The food was not virtuous, and the weight came off anyway. Nobody needs a list of "clean" foods for the energy books to balance.

It does not show that food quality is irrelevant. This was one person over two months, keeping his existing exercise routine, starting from his own baseline. Haub refused to endorse it, telling reporters he was not geared to say it was a good thing to do because there was not enough information. His starting cholesterol was high and he cut his intake sharply, and cutting intake improves those markers in many people whatever the food. A short experiment on one motivated professor with a class watching says nothing about what a person can sustain for a year, what it does to muscle, or what it does to disease risk over decades. His number was his, arrived at for his body and his experiment. It is not a target for anybody reading this.

Why food quality decides whether the deficit survives

Food quality does not change the law. It changes whether you can live inside a calorie deficit without being miserable, and that is the variable that decides most outcomes.

The clearest evidence comes from studies where people were free to eat as much as they wanted. In a controlled feeding trial run at the NIH Clinical Center and published in Cell Metabolism, 20 adults spent two weeks on an ultra-processed diet and two weeks on an unprocessed one, in random order. The meals were matched for calories presented, energy density, protein, carbohydrate, fat, sugar, sodium and fibre. Participants were told to eat as much or as little as they liked. On the ultra-processed diet they ate about 508 kilocalories a day more, and gained an average of 0.9 kilograms. On the unprocessed diet they lost about 0.9 kilograms. The energy was there in both cases. Only one of them made people stop eating.

Protein is the single most reliable lever here. In a University of Washington study published in the American Journal of Clinical Nutrition, 19 participants moved from a diet with 15 per cent of energy from protein to one with 30 per cent, with carbohydrate held constant. Satiety rose markedly. When they were then allowed to eat freely on the higher-protein diet, their spontaneous intake fell by an average of 441 kilocalories a day and they lost weight over 12 weeks without being asked to restrict anything. Westerterp's review notes the same thing from the other direction: protein has the highest thermic effect of the macronutrients and plays a key role in satiety. If you read one thing after this, make it protein for weight loss.

Volume and fibre matter too. In a satiety index built at the University of Sydney, 38 foods were served in equal-energy portions and satiety was rated over two hours. The spread was enormous. Boiled potato scored roughly seven times higher than a croissant for the same energy. Across the whole set, protein, fibre and water content were positively associated with fullness, and fat content negatively. Heavier, wetter, more fibrous food fills you up for fewer calories. That is not a moral claim about food. It is a practical one about how far your allowance stretches.

So the question to ask of any meal is not "is this healthy" but "how will I feel four hours after eating this". A meal built around dal, vegetables, curd and a sensible amount of rice or roti answers that better than the same energy in biscuits, and the difference compounds across every day of a six-month effort.

Do you have to count calories to lose weight?

No, and plenty of people never do. Counting is one tool for getting information, and it is a useful one for a few weeks because most people genuinely do not know what they are eating until they look. But it is not the mechanism, it is a measurement, and it comes with real costs: it can be tedious, it can be inaccurate for home-cooked food, and for some people it turns eating into an anxious accounting exercise, which is a good reason to stop. The alternatives work through the same physics by other means: fixing a rough portion size for rice or roti, putting protein and vegetables in every meal, cutting drinks that carry energy without filling you, eating at roughly regular times, and not keeping the food you cannot stop eating in the house. If you find that tracking food makes you preoccupied or anxious, that is a signal to work with a professional rather than to try harder.

How to arrive at your own number

Start with an estimate, then let reality correct it. That is the whole method, and it beats any equation used on its own.

An equation such as Mifflin-St Jeor or Harris-Benedict predicts basal metabolic rate from height, weight, age and sex. Our BMR calculator does that arithmetic. Multiply by an activity factor and you get an estimate of total daily expenditure, which is what the TDEE calculator produces and the calorie calculator turns into a starting intake range.

Be clear about what those numbers are. They are population averages applied to you. Roughly a quarter of the between-person variation in basal metabolic rate went unexplained in the Johnstone study, and NEAT varies enormously between people who look alike, so your true maintenance level could sit well above or below what any calculator says. The estimate is a place to start, not a fact about your body.

The real number comes from what happens next. Eat at the estimate, as consistently as you can, and record weight and food for two to three weeks. Average the daily weights across each week, because water shifts swamp fat changes over short periods. Then read the trend, not the day. If weight is stable, that intake is close to your actual maintenance level, and you now know something the calculator only guessed. If it is drifting, adjust in a small step and watch for another two weeks. Repeat. It is unglamorous, and it is the only method that accounts for the parts of your physiology nobody can measure from a form. A nutrition professional can run this with you, and doing it well is a clinical skill rather than a piece of arithmetic.

Why does the scale go up on a day you ate less?

Because short-term weight change is mostly not fat. When researchers measured what a two-week weight change is actually made of in free-living adults, using isotope dilution and DXA, a unit of body weight change was 84 per cent fat-free mass, and it tracked changes in total body water. A large part of that is glycogen, the body's stored carbohydrate, which is held in hydrated form with three to four parts water, so eating more or less carbohydrate moves the scale without touching your fat. Add a salty meal, a hard workout, constipation, poor sleep or the days before a period, and the scale can move a kilogram in either direction with no fat involved. Real fat changes are slow and small by comparison, which is why a single morning's reading carries almost no information and a fortnight of readings carries quite a lot. If you weigh at all, weigh at the same time under the same conditions, average the week, and judge the line rather than the point. If daily weighing makes you anxious, weigh weekly or stop weighing and use how your clothes fit instead, because a measurement that damages your relationship with food has a negative value however precise it is.

How fast, and why slower usually wins

Slower is not a moral position. It is a practical one, and there are three concrete reasons for it.

The first is muscle. Norwegian researchers randomised 24 elite athletes to lose weight at either 0.7 per cent or 1.4 per cent of body weight per week, with resistance training in both groups, and published the results in the International Journal of Sport Nutrition and Exercise Metabolism. Both groups lost a similar total amount of weight. The slower group increased lean body mass by about 2 per cent; the faster group's lean mass did not change. Same weight off the scale, different composition underneath.

The second is adherence. A very aggressive calorie deficit is an unpleasant place to live. Hunger rises, energy falls, social eating becomes a problem, and the plan usually breaks before the goal arrives. A deficit you abandon in week five loses to a smaller one you hold for six months, every time.

The third is the rebound. The Biggest Loser follow-up is the extreme illustration: extraordinary loss under extraordinary conditions, followed by regain of most of it and a resting metabolic rate still hundreds of kilocalories a day below baseline six years later. That study cannot tell us how much of the metabolic slowing was caused by the speed of the loss specifically, and it should not be read as proof that fast loss ruins your metabolism. It should be read as evidence that the body defends its weight for a long time, and that a plan with no life after the diet is not a plan.

For anyone whose weight is affecting their health, the sensible route is a plan built with a professional rather than a rate copied off the internet. Our guide to managing obesity in India covers what that assessment involves. If a drug class comes up in that conversation, remember that weight-loss medication is prescription-only and a doctor's decision, not a consumer choice.

The Indian kitchen problem

Most calorie advice online assumes a packaged, portioned, labelled food supply. Indian home cooking is none of those things, and this creates real, specific problems.

The first is shared oil. A sabzi is cooked once for the whole family in a single kadhai. The oil goes in by eye, and the dish is served by ladle. Nobody knows how much of that oil ended up on any one plate. The ICMR-NIN Dietary Guidelines for Indians, 2024 separate dietary fat into invisible fat, which is present in the food itself, and visible or added fat, meaning cooking oils, ghee and butter. The guidelines classify a diet as high in fat when added oils and fats exceed 30 grams a day at a 2,000 kilocalorie level, and they name deep-fried and oil-heavy items directly: samosa, kachori, puri, parathas, savouries, biscuits, cakes, and even some curries. Oil is the biggest single uncounted variable in an Indian kitchen, and it is also the easiest to reduce without changing what you eat.

The second is the absence of portions. There is no label on a bowl of dal. Roti sizes vary by household, by cook and by mood. Rice is served by whoever is serving. This is why our Indian food calorie chart is built around household measures like one katori, one medium roti and one ladle, rather than grams that nobody weighs. An approximate number you will actually use beats a precise one you will not.

The third is food eaten outside. Restaurant, canteen and street food is cooked with more oil, sugar and salt than home food, and served in portions you did not choose. There is no reliable way to estimate it from a menu. So do not calculate it, plan around it: decide in advance how many such meals a week you are having, keep the rest of the week simple and protein-forward, and stop trying to reconstruct the arithmetic afterwards. Our Indian diet plan for weight loss is built on that assumption rather than on the fiction of a fully controlled week.

The fourth is the family meal. Most people in India do not cook only for themselves, and a plan that needs separate food usually collapses within a month. Adjusting the shared dish, the oil in it and the size of your own serving lasts longer than cooking twice.

What to take from all of this

The law is not the problem. A calorie deficit is what produces weight loss, and nothing else does. The problem is that "calories in, calories out" is presented as an instruction when it is really a description, and the two numbers inside it are estimates with wide error bars that move while you use them.

So use the arithmetic for what it is good for. It tells you the direction to go and roughly how far. Then get the real number from your own body over a few weeks, choose food that leaves you full enough to keep going, protect your muscle with protein and resistance work, go slowly enough that you can still be doing it in six months, and stop treating a single week's scale reading as a verdict on your character.

If this is the kind of thinking you want to do properly, and for other people rather than only yourself, it is what a formal qualification teaches. NNWA's Diploma in Nutrition, Dietetics & Public Health covers energy balance, dietary assessment and counselling over six months, taught online in English and Hindi by practising professionals. One honest note about what that is and is not: a skill qualification is not a degree, and Registered Dietitian status in India requires a BSc or MSc plus registration with the Indian Dietetic Association.