To those folks, I suggest they read the comments and the social media commentary, and they will find many people with the second objection:
2. Conservation of energy doesn’t apply to the human body! It’s hormones or mass or something, and calories are just an invention of the processed-food industry.
To those folks, I suggest they talk to a physicist.
And now, the interesting argument:
3. Yes, the calories-in/calories-out math has to work, but low-carb diets affect both sides of the equation, so you can’t dismiss (as I did) the effect as real but small.
Everyone agrees that the macronutrient content of your diet can affect how many calories you burn (because fat, carbs, and protein are metabolized differently) and how many you ingest (because some foods are more satiating than others). The disagreement is over the magnitude of those changes, particularly with regard to low-carb diets. So let’s talk about that.
First is that low-carb diets affect the calories-out part of the equation. Fewer carbs mean less insulin, and because insulin is integral to fat storage, the reasoning goes, you end up burning calories instead of storing them. That’s a plausible, testable theory! Bring people into a controlled setting, and feed some people low-carb diets and others higher-carb diets with the same number of calories, and see who loses more weight.
Then there’s the calories-in side of the equation. Because low-carb diets are more satiating, the reasoning goes, it’s easier to eat fewer calories. That’s also a plausible, testable theory! Assign people different kinds of diets without caloric restrictions and see who loses more weight.
Lucky for us, both of those tests have been done, so let’s look at the evidence.
Ah, evidence. It’s a sticky wicket, because you can find “evidence” to support just about any conclusion you would like to draw. And, if you traffic in diet discourse, you’ve no doubt seen long lists of evidence that support the low-carb theory.
When you look at a list like that, here’s the question you should ask: Is someone trying to prove that something is true, or discover whether it’s true? As a journalist, my job is B, but I’m as susceptible to confirmation bias as the next guy, so I’ll do my level best to look at all the evidence. I’m sure you’ll keep me honest.
Let’s take calories-in first. If you really want to discover whether you eat less on low-carb diets, park yourself over at PubMed, the repository of journal articles. I focus on meta-analyses, because their job is to assess the preponderance of the evidence. If you use “low-carbohydrate diet weight loss” as your search terms, and restrict the results to meta-analyses, you get 61 results. I read every one that seemed relevant.
They all said the same thing, which is comforting because they’re mostly meta-analyzing combinations of the same set of studies. The net is that, in short-term trials (a year or less), low-carb diets outperform other diets by a few pounds (under 10). Examples include Chawla, 2020; Monsoor, 2016; and Nordmann, 2006. Two meta-analyses, though, find that people lose a bit more weight on the Mediterranean diet than low-carb. (Ajala, 2013; Pan, 2018).
After a year, any low-carb advantage all but disappears. Some meta-analyses find a small advantage, but we’re talking two or three pounds (Bueno, 2013; Tobias, 2015). Others find no advantage at all (Rafiullah, 2022; Silverii, 2022; Naude, 2014; Hu, 2012).
My read is that low-carb diets are easier to follow, short-term. Taking whole categories of calorie-dense, easy-to-eat foods (bread, pasta, rice, baked goods) off the table works for a lot of people, at least for a while. And there’s some evidence that very-low-carb diets (the ketogenic kind) are more satiating than other diets – but also a compelling but short-term study that found keto dieters eating nearly 700 calories more per day than low-fat dieters.
In the long run, though, what matters about a diet is simply whether you can stick to it, and trials indicate that low-carb isn’t better than other diets in that regard.
On to the calories-out part of the equation. Do people on low-carb diets lose more weight on the same number of calories because their bodies burn more energy?
In strictly controlled trials where subjects are fed isocaloric diets of varying macronutrient composition, they don’t. There have been a number of these trials, and I couldn’t find a single one where carb content made a difference.
Back in 1977, a small study pitted a 14-day diet that was 70 percent carbs against one that was 10 percent carbs. It found that subjects did lose more weight on the low-carb diet, but they immediately gained it back when the diet was discontinued, indicating that it was the fluid loss you expect from very low-carb diets. A longer study in 1992 tested diets from 15 percent to 85 percent carbs, and found no difference in the energy required to maintain body weight.
In 2015, National Institutes of Health researcher Kevin Hall found that subjects on a low-carb diet burned less fat (53 grams/day) than those on a low-fat diet (89 grams/day). And in 2016, Hall found that a ketogenic diet did indeed lead to more energy burned but that the small difference didn’t correspond to fat loss.
I wanted to make sure I wasn’t missing anything, so I checked in with David Ludwig, professor of nutrition at Harvard School of Public Health and a prominent low-carb diet advocate. He pointed me (by e-mail) to his meta-analysis, not of weight loss but of total energy expenditure (TEE), which shows that low-carb diets do lead to higher energy expenditure, but the biggest effects don’t happen until after about 17 days.
Hall and Ludwig have disagreed about how to measure TEE and how these studies should be interpreted, and although they see things very differently, I think their back-and-forth is a terrific example of how scientific disagreements should be handled. Hats off to both.
Ludwig hangs his hat on TEE, which, he told me, is an “instantaneous measure” that doesn’t require you to wait until the deficit shows up as fat loss. But, as Hall has written, sometimes TEE doesn’t add up, especially in outpatient studies. One study found that low-carb increases TEE by 200 to 300 calories per day, for example, but reported 480 fewer calories eaten than expended, despite average weight being stable.
I’m gonna go out on a limb and suggest that the gold standard for weight-loss trials is actual weight loss, and if the amount of excess energy burned on low-carb diets were significant, we would expect to see it show up in weight (or, more particularly, fat) loss. And it doesn’t.
Ludwig explained this to me: “Anything less than 3 months is, in my view, so hopeless[ly] confounded by transient effects, and so underpowered, as to be virtually meaningless.”
A three-month study of inpatients, with diets carefully controlled, is probably prohibitively expensive, but I guess that’s where I’ll rest my case. Yes, low-carb diets can indeed increase energy expenditure, but by so little that the signal wouldn’t rise above the noise until at least 3 months.
I am in favor of low-carb diets; they eliminate big groups of foods that most of us should probably eat less of, and some people find a way to stick to them long-term. But there’s no metabolic magic.
You lose weight on low-carb diets because … drumroll, please … you eat less.
