Why Calories Count
July 23, 2013 by Armi Legge 59 Comments
You’ve heard calories are important.
You’ve been told for years that weight loss is about calories in versus calories out.
But you’re not happy with that explanation.
If you want to lose weight, you need to eat fewer calories than you expend. Here’s why.
Image
You’ve heard other people say that calories don’t count — that thermodynamics don’t apply to humans. You’ve heard that weight loss is endlessly complex and that calories are a “scam.”
Both sides say their ideas are supported by scientific evidence and you’re not sure which to believe.
To see which side is right, let’s look at how to set up an experiment to test the idea that calories count when it comes to weight loss. Then we’ll look at the studies that meet these guidelines, the ones that don’t, and what they both show.
How to Conduct a Weight Loss Study
We want to know if eating fewer calories causes weight loss.
Good experimentation is mostly about controlling for confounding variables. In our case, we need to control things that might influence calorie intake and weight loss.
We need some subjects. Since we don’t care about helping rats or mice lose fat — we need humans.
We recruit some slightly overweight or obese people to make the results generalizable to your average person who’s trying to lose weight.
We make sure all of our participants are as similar as possible in terms of weight, health, age, gender, dieting history, etc. This is to make sure these factors don’t distort our results.
After we’ve got our subjects, we begin the experiment.
We need to measure how many calories these people are eating to establish a baseline. We measure their energy expenditure with indirect calorimetry — where the subject’s breath is analyzed for different gases to indicate how many calories they’re burning. This gives us an excellent estimate of their energy expenditure.1,2
To make sure our estimates are correct and adjust their food intake to match their activity levels, we feed them a specific amount of calories for a week or two and monitor their weight. We adjust their calorie intake as needed so they’re eating the same amount they’re burning until they maintain their weight. This is referred to as an “eucaloric diet,” or maintenance.
We prepare all of our subjects’ meals ahead of time. We weigh every gram of food with a digital scale and we watch them during meals to make sure they eat every scrap of food. We also feed all of our subjects the same foods.
We give our subjects instructions to maintain about the same level of activity throughout the study.
Once we find how many calories each of our subjects needs to maintain their weight, we can take our experiment in several directions.
1. To test if eating less causes weight loss…
We would randomly assign half of our subject to a group that eats fewer calories every day. We would assign the other half of our subjects to a group that keeps eating their normal amount of food. If the group that eats less loses weight, then it’s likely that calories count.
2. To test if different macronutrients affect weight loss…*
We would split the subjects into two or more groups, feed them different macronutrient ratios, and put them on the same caloric intake. If one group loses more weight than another, it’s possible that whatever they were eating accelerated (or hindered) their weight loss.
3. To test if different people lose more or less weight on different diets…*
In random order, we would make all of the subjects eat several different diets. We would include a “wash-out” period in between each diet to get a new baseline for each subject. If the subjects consistently lost more weight eating a certain balance of protein, carbs, and fat, it’s possible that a specific macronutrient ratio accelerates weight loss.
This “cross-over” type experiment is also good if you have a small number of subjects (which many weight loss studies do).
4. To test if eating more calories causes weight gain…*
We would make one group eat above their calorie needs, and the other group eat at maintenance. If the former group gained weight, it’s likely that calories count.
These are the main kinds of controlled weight loss experiments you usually see, but there are endless possible variations. We could (and eventually should) also include studies with exercise.
The key is to make sure our subjects do what we tell them.
We want to know for a fact that these people are eating exactly as many calories as we tell them to.
We have to keep them locked in the hospital where we can monitor all of their activity and food intake (no cheating).
We need to keep weighing all of their food and watching them at every meal.
We keep them on these diets for several weeks or months, and measure their weight. If we’re really trying to do a great job, we’d also measure their body composition, energy expenditure, and maybe a few other variables as well.
Like any good study, we’re sure to blind the researchers (us and our colleagues) and the subjects as to who is getting which treatment.
We would need to repeat this study on other people of different weights, genders, ages, and ethnic groups. We would test people with different diseases and different dieting histories (e.g. always been overweight, gained weight recently, etc.). We would also conduct studies that involved exercise and see how it interacts with the diet.
Then we would share our results and wait for other independent researchers across the globe to repeat them — using the same study design.
If other scientists found similar results to our studies, on similar subjects, using the same or similar conditions, we could assume with a reasonable level of certainty that our findings were correct. Then we’d throw a crazy party and blow a bunch of funding money (kidding).
If our findings were repeated consistently for years — even decades — and we were able to logically explain any odd results, we could call our theory a fact (an idea with the highest level of certainty).
Weight loss researchers have done exactly this (except for the parties) for over a century. Let’s see what they’ve found.
When People Create a Caloric Deficit — They Always Lose Weight
Studies using the rigorous standards outlined above have consistently shown that when people eat a calorie restricted diet — they lose weight.3-6
That is, when researchers measure people’s energy expenditure, weigh all of their food and count their calorie intake, and force them to eat less food than they need to maintain their weight — they lose weight. The amount of weight they lose is also generally proportional to the size of their caloric deficit.
Whether they eat mostly protein, fat, or carbs makes no significant difference in how much weight they lose.3,5,7-9 Eating more protein does cause you to burn slightly more calories, around 70-100 per day, but it’s usually not enough to make a significant difference in weight loss.10-13
At this point, eating a calorie restricted diet and exercising (largely to burn more calories) are also considered the two most scientifically supported ways to lose weight.14-17 It works.
When you restrict your calorie intake enough to create a deficit, you lose weight. It turns out the reverse is also true.
When People Eat More Calories than They Need — They Always Gain Weight
In studies where subjects are forced to overeat — they always gain weight.18-24 25-28
There is often a lot of individual variation in how much weight and fat people gain, but they always gain some. These differences are mostly because some people subconsciously move more when they overeat to burn off the extra calories (a phenomenon called non-exercise activity thermogenesis, or “NEAT,” which we’ll talk about later).19,24,29-32
However, this increase in subconscious energy expenditure is never enough to completely offset the increase in calories. In every study thus far — people still gain weight when they are eating more calories than they expend. They never expend enough to completely prevent weight gain when they overeat calories.
On the other hand, not all studies seem to show that eating fewer calories is necessary for weight loss, or that eating more calories causes weight gain.
Why Some Studies Seem to Show that Calories Don’t Count
As usual, people claim that you can find a study to prove anything. You can’t, but here’s why people think that’s true when it comes to calories and weight loss.
Studies often report that people eating low-carb diets lose as much or more weight as those eating high-carb diets, sometimes despite the subjects claiming to eat the same or more calories.33-45
There are also a similar number of these studies that have found greater or no difference in weight loss between high- and low-carb diets, but like most people who claim calories don’t count, we’re going to ignore those for now (and come back to them in a minute).46-62
Based on these studies people claim that:
1. You can lose weight without a caloric deficit.
2. You can lose more weight on the same calorie deficit by avoiding certain foods or macronutrients, like carbs. Thus macronutrients are what you should focus on — not calories.
Both of these claims are completely untrue.
None of these studies controlled for the variables we talked about at the beginning of this article. Most of these studies were conducted under “free-living conditions,” which means the subjects were given instructions on what to eat, sent home, and told to track their food intake.
They chose what — and how much — to eat and the researchers had to trust their diet records.
This is a big problem.
Diets Help People Eat Less Without Realizing It
When people go on a diet, they tend to spontaneously eat less for several reasons.
First, diets place restrictions on what and when you can eat.
No carbs.
Low-fat.
No meat.
No gluten.
No grains, dairy, or legumes.
No eating late at night.
Eat every three hours.
Don’t eat breakfast.
Intermittently fast.
No processed foods.
No sweets.
No sugar.
When you take away a signifiant portion of someone’s diet or place limits on when they can eat, they can’t help but eat less (at least for a while). If someone’s used to getting 50% of their calories from carbs — and they cut that to less than 10% — they’re going to eat less until they get used to their new diet.
Placing a limit on food variety also tends to make people eat less without realizing it. If you have fewer food options, you get bored faster and tend to eat less.63 When people have more options, they eat more.63,64
Perhaps the biggest reason people eating low-carb diets sometimes lose more weight is that they also tend to eat more protein. Studies have repeatedly found that when people eat more protein, they feel fuller and spontaneously eat fewer calories without knowing it.65-70
When people switch to a high protein low-carb diet, they sometimes spontaneously reduce their food intake by 1,000 calories per day.71 These people didn’t notice the difference, yet in just two weeks they lost 3.6 pounds (1.65 kilograms).
You’d think people would notice if they were eating significantly fewer calories. They don’t.
People Have No Clue How Much they Eat
People are horrible at estimating their calorie intake.72-101 102-120
Overweight and obese people (especially women) are often the worst, but most people underestimate their calorie intake to some degree.
It’s true for men and women and people of all ages.
It’s true when people are given specific instructions on how to measure their food intake.
It’s true for dietitians.102
It’s true even when people are paid to track their food intake.104
In some cases, people who claim they can’t lose weight by cutting calories underestimate their food intake by 47%, and overestimate their exercise levels by 51%.75 Other data has shown that people can underreport their food intake by up to 2,000 calories per day.89
It’s likely that the people on high carb diets are more likely to underreport their food intake. This would make it seem as if people on low-carb diets are losing weight despite eating more calories.
The people on high carb diets in weight loss studies are often told to consciously restrict their calorie intake and avoid “unhealthy” or “bad” foods, especially fat. These are all behaviors that generally increase the likelihood that people will underreport their food intake.81,92,97-101,121
On the other hand, the people eating the low-carb diets are often told to eat as much fat and protein as they want.
Remember that most of these people were probably making some attempt to control fat intake before the study, or were at least used to the idea that fat is “bad” (thanks largely to the USDA, FDA, and other health agencies). When they’re told to eat a low-carb high-fat diet and to eat as much of these previously “forbidden” foods as they want, even small amounts can feel like a lot of food.
People on low-carb diets often eat more total protein and fat, which helps blunt their appetite.65-70,122 In contrast, the people in the high-carb groups in these studies are often eating lots of refined carbs which tend to be far less filling. In fact, studies have shown that it’s actually the high protein content of the diet that helps control appetite and cause weight loss, not the avoidance of carbs.123
Enjoying previously taboo foods, eating more protein and fat, and not being told to restrict calories drives people in the low-carb groups to eat less and report they’re eating more than they really are.
This effect wears off, however. These people generally get used to their new diet and start eating more of the low-carb high-fat foods —* and thus total calories. Over time they also tend to get bored with their diet and become less compliant.
This is why most free-living studies lasting longer than six months have found that people on high- or low-carb diets lose the same amount of weight.33,43,45,62,124,125 It’s probably also why many free-living studies have found that people lose the same amount of weight eating high- or low-carb diets.
This is why you should be highly skeptical of people who claim they lost weight without eating fewer calories. Calories count. These people are just not counting them accurately, if at all.
If eating a low-carb diet helps you lose weight without counting calories — that’s great — as long as you can maintain it or find another option if it stops working. However, that doesn’t mean calories don’t count or that low-carb is the ONE TRUE DIET™ that will work for everyone.
The people who lost more weight in these free-living studies were eating fewer calories — they just didn’t know it.
Low-Carb Diets Make People Lose More Water Weight
There are several controlled studies that have found people eating low-carb diets lost more weight than those on high-carb diets.126-129 Unfortunately, these studies have other problems.
While the people on low-carb diets lost more weight, they didn’t lose more fat — they lost water weight.128,130
When you eat carbs, they’re largely converted into muscle glycogen — the storage form of carbohydrate. Every gram of glycogen is bound to about 3-4 grams of water.131,132
When you switch to a low-carb diet, your glycogen levels drop, and you lose water weight.133-136 Low-carb diets also tend to deplete electrolyte levels which can also have a diuretic (water depleting) effect.
Several of the studies that found low-carb diets helped people lose more weight only measured total body weight. They didn’t measure body composition, so there’s no way to know whether they lost fat, water, or muscle.
Other studies have measured or controlled for changes in water weight and have found no difference in fat loss.
Granted, it’s fun to see big changes in scale weight when you slash carbs, but those aren’t necessarily big changes in fat loss.
You Are Not a Rodent
Rodents also only lose weight when they’re in a caloric deficit, but there is at least one rodent study that found a ketogenic diet (generally less than 100 grams of carbs per day in humans) increased metabolic rate and weight loss more than a high-carb diet.137
However, there are several reasons these results are far less important than they appear at first sight.
First of all, there are other studies showing that rats overfed on a ketogenic diet still gain significant amounts of fat — more so than on high-carb diets.138 In another mouse study where a ketogenic diet also increased energy expenditure and prevented weight gain, it still caused insulin resistance.139
Other studies have also found that rats placed on a low-carb diet don’t lose weight unless they also eat fewer calories.140
However, none of these animal studies show how a specific diet will help you lose weight. Rats are not humans, and studies on rats, mice, and other animals can never be completely generalized to people.
Rodents have several different physiological traits that may make it easier for them to gain fat on high-carb diets.141-143 They convert carbohydrate to fat (a process called “de novo lipogenesis”) about 10 times more efficiently than humans do.144,145 This is true for other animals like pigs as well.146 Rodents also process protein and fat differently.147,148
It’s fine to make educated guesses based on rodent studies of how something might work in humans. However, those results have to be validated by controlled human studies before you can call them anything more than “interesting.”
If we later find that the same diet that caused fat loss in rats didn’t work for humans, and this holds true for multiple studies, then it doesn’t work for humans — period. That’s exactly what we’ve found with low-carb diets and weight loss.
How to lose weight is one of the most thoroughly researched questions in the past 100 years, and at this point we have enough good human trials to stop worrying about rat studies. Not that they don’t have some value, but they don’t meet the same standard of evidence that human studies do.
Calories Count
