CICO Science: Does Calories In vs Calories Out Really Work for Fat Loss?

Does CICO work? As physics, yes. Body weight changes follow the balance between energy in and energy out, and no hormone, food group or fasting window has been shown to switch that off. But the “calories in, calories out” slogan hides a lot: both sides of the equation move when you diet, and neither is easy to measure.

This article defends the science of CICO and explains where the slogan falls short. It sits inside our wider guide to energy balance for fat loss, bulking and recomposition, and it is the foundation for every cutting or bulking decision on this site.

What does “calories in, calories out” actually claim?

Two things, and people constantly mix them up.

The first is a law of physics. Energy is not created or destroyed. The energy in food is either used for heat and movement, stored as fat, muscle or glycogen, or lost as waste. Over any stretch of time, if more energy goes in than gets used, the difference is stored. If less goes in, the body draws on its stores.

The second is a theory of why people gain weight. That is where most of the arguments start. A 2022 paper by Hall and colleagues makes this exact point: the principle of energy balance is a law of physics that is agnostic as to obesity mechanisms, while the “energy balance model” of obesity is a biological model in which the brain controls food intake through endocrine, metabolic and nervous system signals.

So when someone says “CICO is wrong because hormones matter”, they are usually attacking a cartoon version: the idea that weight is simply willpower and a calculator. Few researchers in the field hold that view.

Is the science of CICO real? What controlled studies show

When researchers lock people in a metabolic ward and control every meal, the picture is consistent: weight follows energy balance.

A 2019 NIH trial by Hall and colleagues is a good example. Twenty weight-stable adults (mean BMI 27) lived in a clinical centre and ate an ultra-processed diet for 2 weeks and an unprocessed diet for 2 weeks, in random order, eating as much or as little as they liked. Meals were matched for presented calories, energy density, macronutrients, sugar, sodium and fibre. Intake was about 508 kcal per day higher on the ultra-processed diet. Participants gained 0.9 kg on it and lost 0.9 kg on the unprocessed diet. Weight change was strongly correlated with energy intake (r = 0.8).

Read that carefully, because it supports CICO and complicates it at the same time. Weight tracked intake closely. But the food changed how much people ate without them trying. That is a story about appetite (the “in” side), not a broken equation.

This was a small, short trial of 20 adults, so treat the numbers as an illustration rather than a prediction for your own diet.

Do carbs, insulin or the keto diet get around thermodynamics?

This is the biggest objection, usually phrased as “insulin makes you store fat, so calories do not matter.”

The theory behind it is called the carbohydrate-insulin model. It predicts that if you swap carbohydrate for fat at the same calories, insulin falls, energy expenditure rises and you lose more body fat. That is a testable claim, and it has been tested.

A metabolic ward trial. In a 2016 study, 17 overweight or obese men ate a high-carbohydrate baseline diet for 4 weeks, then an isocaloric ketogenic diet for 4 weeks with protein held constant. Energy expenditure did rise slightly on the ketogenic diet (for example, 151 kcal per day by doubly labelled water). But body fat loss slowed, and the authors concluded the diet was not accompanied by increased fat loss, with the expenditure increases near the limits of detection.

A meta-analysis. A 2017 analysis of 32 controlled feeding studies with isocaloric swaps of carbohydrate for fat found that both energy expenditure (26 kcal per day) and fat loss (16 g per day) were greater with lower-fat diets, not lower-carb diets.

Why do findings differ at all? Studies use different durations, protein levels and ways of measuring expenditure. Small differences of 50-150 kcal per day are hard to detect and sit at the edge of measurement error. The cautious reading is this: low-carb diets can help some people eat less because they feel full or cut out snack foods, but there is no good evidence of a large metabolic advantage at equal calories.

Insulin is not irrelevant. It is part of the signalling system that governs fuel use and hunger. It just has not been shown to override energy balance. If you want the lifter angle, our piece on low-carb and ketogenic diets for hypertrophy goes deeper.

Where do thyroid hormone and other hormones fit in?

Hormones change CICO by pushing on one side of it.

Thyroid hormone is the classic example. It influences resting energy expenditure, so an underactive thyroid can lower the “out” side a bit. But how much? A 2026 systematic review and meta-analysis by Wolde Sellasie and colleagues pooled 23 studies (763 patients) of hypothyroid adults treated with levothyroxine until their thyroid levels normalised. Body weight fell by an average of 1.39 kg (95% CI: 0.23 to 2.56 kg), with reductions seen in overt hypothyroidism but not subclinical hypothyroidism. The authors described the effect as modest and discouraged levothyroxine solely for weight loss.

So an untreated thyroid problem is not a magic reason you cannot lose weight, and a normal thyroid is not a magic way to lose it. The effect is small compared with the swing created by what you eat and how much you move.

The same logic applies to cortisol, leptin and appetite hormones: they shift hunger and energy use. They do not cancel out a deficit.

Why does calorie counting fail so often?

If the physics is solid, why do so many people say counting “does not work”? Mostly because of the measurement, not the law.

Intake is under-reported. A review of studies using doubly labelled water (an objective measure of energy expenditure) found that when people record their own intake, it is generally under-reported compared with measured expenditure. The review noted that other factors beyond body size, such as dietary restraint, are involved. The in-gym version of this is the thing people forget: the tadka oil, the handful of peanuts, the extra spoon of ghee on roti. Our guide to hidden calories in Indian cooking covers where these hide.

Expenditure moves when you diet. In a 2011 Lancet modelling paper, Hall and colleagues showed that bodyweight response to a change in intake is slow, with half times of about 1 year, because expenditure adapts as weight changes. So a fixed “500 kcal deficit equals 0.5 kg per week forever” is not how it plays out. The model also found that a small persistent imbalance of about 30 kJ per day (roughly 7 kcal) underlies average weight gain at population level, which shows how small errors in tracking can matter over time.

Metabolic adaptation is real but not magic. A review of metabolic adaptation in athletes describes how energy restriction is accompanied by changes in hormones, mitochondrial efficiency and energy expenditure that minimise the deficit. That makes the deficit smaller than the spreadsheet says. It does not flip it into a surplus. We cover the details in our article on metabolic adaptation and why starvation mode is mostly a myth.

How does NEAT change the “calories out” side?

Not all of expenditure is under your control, and NEAT is the clearest example.

In a 1999 Science study by Levine and colleagues, 16 non-obese volunteers were fed 1,000 kcal per day above maintenance for 8 weeks. Two-thirds of the rise in daily energy expenditure came from NEAT (fidgeting, posture, daily activity). Changes in NEAT accounted for a 10-fold difference in fat storage between people, and predicted resistance to fat gain (r = 0.77).

Flip it around for a cut. When you eat less, daily movement can drift down without you noticing: fewer trips, more sitting, less fidgeting. That is one reason a deficit can shrink on its own. The study was in non-obese adults overeating, so the cutting application is an inference. Still, a simple step count is a cheap guard. Our guide to TDEE, BMR, TEF and NEAT breaks down each component.

So how do you use CICO properly?

Think of it as a feedback loop, not a formula.

1. Get a starting estimate. Use the calorie calculator to estimate maintenance. Treat it as a first guess with a margin of error.
2. Log accurately for 2 weeks. Use a digital scale for calorie-dense items such as oil, ghee, nuts, peanut butter, paneer and rice. See food logging precision.
3. Track a weekly average. Weigh daily if it does not stress you, then compare weekly means.
4. Adjust in small steps. If the trend is flat after 2-3 weeks, trim 100-200 kcal or add steps. Do not slash intake by half.
5. Keep the food simple. Dal, eggs, curd, paneer, soya, chicken, vegetables, roti and rice are filling and easy to portion.

Situation	Likely side of the equation	First thing to check
Weight flat, logging looks perfect	Calories in (under-reporting)	Weigh oil, ghee, snacks, weekend meals
Losing fast then stalling	Calories out (adaptation, lower NEAT)	Weekly step average, recalculate maintenance
Hungry all the time	Calories in (appetite)	Protein, fibre, volume of food, sleep
Tired, cold, hair changes plus flat weight	Possible medical issue	Talk to a doctor

For a longer discussion of plateaus, see breaking fat loss plateaus.

The bottom line

CICO is true the way gravity is true. It does not tell you how to stay on a diet, how hungry you will be or what your expenditure will do. But it does rule out the claim that any single hormone or macronutrient lets you lose fat while eating more energy than you use.

The practical version is less dramatic. Eat in a modest deficit, pick foods that keep you full, protect your daily movement, lift to keep muscle, and judge progress on trends. If your own data stay stubborn after careful tracking, see a doctor instead of trusting a forum thread.