On Weight Loss and Weight Loss Drugs
An elaboration on some tweets.
Some days ago I twote the following:
the "just eat less" crowd do not have an explanation for how obesity emerges (let alone how to resolve it on a population scale), because it emerges from approximately a 10 calorie per day "surplus"
This was a controversial tweet that made a lot of people angry, and I argued for a solid couple of days after, as part of a broader defense of weight loss drugs.
However, in hindsight, I fear a lot of people didn’t read this tweet correctly (as well as my various responses), and, seeing as I haven’t written much recently, I figured I might as well write about weight loss.
Calories In, Calories Out?
The “traditional” model of why weight gain happens is that people who gain weight are doing so because they’re eating more calories than they’re burning. To legions of fitness influencers and their followers, this seems very obvious. After all, if you’re overweight, it’s because you’re gluttonous for greedily eating more calories than you’re burning, and if you would just put the fork down, you would lose weight. People who fail to adhere to a diet are just of weak constitution, and should just try harder.
Now, there is one thing I will give the fitness bros — it is true that if your body admits more usable energy than it burns, it will gain weight. That is the first law of thermodynamics, and the first law of thermodynamics isn’t somehow completely wrong in the case of human body weight.
However, if we follow this line of argumentation, we might ask how the theory addresses one of the fundamental questions of obesity: why did it start appearing around 1970?
The assumption behind the “calories in, calories out” thesis supposes that, around 1970, we started to experience a marked decline in our willpower, and as a result ate more calories — after all, there is a higher supply of calories per person today than in the 1950s and 1960s, and presumably a lot of this supply is consumed.
I find the implicit thesis quite implausible. It’s not like Americans just suddenly became weak-willed around 1970 — and increased calorie consumption could be explained by the fact that people weigh more, and heavier people need to eat more to maintain their basic bodily functions.
Even the more sophisticated version (that calories became cheaper, and we began to eat more highly-processed food) falls flat, in my estimation — though, I believed this thesis in the past. Unhealthy, highly-processed food was quite available in America before 1970, and countries such as Japan that have markedly low obesity rates still have lots of highly-processed food.
On an individual level, the “calories in, calories out” thesis doesn’t engage with the fact that “calories out” is nigh immeasurable and dependent on calorie consumption. As a result, it’s kind of a bad way to go about dieting.
Energy is burned in our bodies through four primary channels:
thermic effect of food (your body expending energy to break down carbohydrates and proteins; about 10% of calories burned),
exercise activity thermogenesis (what your smartwatch would track; also about 10% of calories burned),
non-exercise activity thermogenesis, or NEAT (basically just all the other movements you do throughout your life, including ones not really under your conscious control like fidgeting; about 10–20% of calories burned),
and basal metabolic rate (the energy needed to keep all your systems at their minimal capacity, more than 60% of calories burned).
Notice that many of the calories your body burns are not through things that are easily measured; it’s not like your smartwatch calorie tracker will measure how many calories you burned through fidgeting today, whether your body is running hot vs. cold, etc, but that’s most of your calorie burn. More importantly, your body can upregulate or downregulate some of these depending on other factors throughout the day. Think about how, if you have gone on a run or intense bike ride, you probably walk up stairs slower and fidget less afterward, or if you eat a very meat-heavy meal, your body starts to run hot.
It’s not much better on the “calories in” front, either. FDA calorie labels can be off by up to 20% (in either direction), and calorie content of a food item is primarily determined through incineration rather than its interactions inside the body. For instance, fibrous foods will often have lower effective calories than their listed value — a recent example found almonds to actually have 20% fewer calories than their previous value.
Overall, this evidence suggests that “calories in, calories out” is hard to measure — and while I’m sure many people can end up benefitting from calorie counting, it is obviously not a scalable solution to obesity.
Outcomes vs. Mechanisms
I don’t dispute the first law of thermodynamics — clearly, if someone is losing weight, it is because they are using more calories than they are burning. However, given the difficulties with calorie counting (on both ends), I think it’s fair to say that “burning more calories than one is consuming” is the desired outcome. It isn’t an actionable mechanism for achieving sustained weight loss.
Notice the emphasis on sustained — this is because there is an easy way to lose a lot of weight in a short amount of time: crash dieting. You could restrict yourself to one or two 400 calorie meals a day, and you would lose weight. However, there’s a reason no one seriously considers doing this — it sucks. You will feel tired at first, and with longer periods of time, you would feel extreme lethargy, be unable to sleep at night, and you would probably end up binging at some point.
If you somehow survived the diet, chances are that, if you allowed yourself to eat normally, you would be extremely hungry and regain a lot of the weight you had lost through the diet. Indeed, most people who go on crash diets end up gaining the weight back — up to 95% of people who attempt one!1
However, something interesting crops up if we look at overfeeding studies — the anonymous internet sleuths at Slime Mold Time Mold looked at a case of prisoners in Vermont who were randomly assigned to eat large quantities of food. Of course, participants who ate a lot of food ended up gaining weight — but, in a fascinating twist, once they stopped their diet, they rapidly lost it and returned to around their original weight.
If we took the “calories in, calories out” idea at its word, we would say that people in overfeeding studies gained weight because they ate more calories than they burned, and they lost that weight because they burned more calories than they ate. However, it doesn’t explain why they lost their excess weight and returned to their original weight. The overfeeding studies also don’t fully comport with the “thrifty gene” thesis — that we are evolved to hold onto weight because of our evolution having gone through periods of famine. If that were the case, why wouldn’t people in overfeeding studies stick to their new, higher weight once they were no longer fed those extremely high-calorie diets?
Set Weights
The example of both overfeeding studies and crash diets suggests that our bodies will defend a certain range of weights that it views as “normal,” through hunger signaling and other mechanisms. This is broadly known as “set point theory,” and does provide a strong explanation for
why people gain, lose, or maintain weight
why diets often fail (but can succeed)
what can be done.
Tackling (1), set point theory states that the reason people gain or lose weight is to return to their set point (or maintain a set range) — and suggests that our bodies will use various mechanisms to ramp up calorie expenditure (such as fidgeting more or less) or tamp down calorie intake (by making us feel satiated or temporarily sick of food), in order to keep calorie intake within a small window. In the case of the tweet, I suggest that there is something subtly off about the set weight — obesity emerges over long time scales, suggesting that, in someone who ends up obese, their set weight slowly rises as their body “chases” after it
To test the idea of a set weight for yourself, on Monday you could eat an extra 500 calories at lunch (I’m fond of goat cheese), and then eat ad libitum for the rest of your week, tracking weight throughout.2 Set point theory suggests that it would change very little if at all once the week is over — and I’d be willing to bet a good amount of money that this would bear out in your n=1 experiment.
When it comes to diets, set point theory can explain failure quite easily: people who attempt a diet will leave their body’s set range, and the body will fight back. If dieting down, they’ll feel hungry, lethargic, unable to sleep, etc. up until the necessary amount of food is eaten to return to the body’s set range, and if attempting to gain weight, they will feel sick of food, exercise, run hot, and fidget more. The easiest example would be the numerous cases of crash dieting that people have attempted.
However, set point theory can also explain success at dieting through two mechanisms. One can be that certain people can have a very wide set range — there are definitely many people who said “hey, I want to lose some weight,” did a diet for a few months, and lost that weight without much effort or the negative experiences of crash dieting. However, it could also be that more gradual dieting, with regular breaks, brings down the set point as your body’s various systems that maintain its set point get settled at the new weight.
Finally, if set point theory is as plausible as is suggested, then the solution for people who are overweight or obese is to reduce their set points. That would imply that their bodies start feeling more satiated, they start moving more, and lose weight without much effort (since their body will defend a new weight that is lower than the current weight).
What this means for you, me, and society
Set weight theory, if it is correct (and certainly, it does seem like the most plausible theory), suggests that we as a society must avoid being so cruel to people who are overweight. It’s not a sign of laziness or gluttony to have a body weight set point at an unhealthy level any more than it is a sign of godliness to have a body weight set point at a healthy level — many factors outside our control end up determining the set point, so why be so mean about it?
When it comes to fitness and health, this suggests that diet and exercise aren’t going to be particularly effective for weight loss — but that should feel liberating rather than discouraging. Instead of tracking calories, people can focus on getting the necessary and proper nutrition to maintain mental clarity and have sufficient energy; instead of worrying about the scale, people in the gym can focus on gaining strength and stamina, all free of self-judgment.
On a more personal level, I grew interested in set point theory after experiencing a long period of depression-induced weight loss. During this period, I just didn’t want to eat anything (and even found a lot of food quite repulsive), and kept losing weight. The depression and its associated weight loss were 100% not of my conscious doing — indeed, people around me were often confounded at my inability to eat normally, and I didn’t have much willpower to force food down my throat. More interestingly, after regaining my appetite, my weight would not budge from the 1–2 pound window I ended up in.
However, accepting set point theory raises a pretty important question — what is to be done for people that have a body weight set point they do not desire?
Enter Ozempic (and other incretin-mimicking drugs)
The weight loss drug du jour — indeed, the synecdoche for weight loss drugs as a whole — is Ozempic (or rather, Wegovy, the weight-loss specific version of the diabetes drug). The active ingredient in Ozempic is semaglutide, which is a glucagon-like peptide-1 receptor agonist — the drug imitates the hormone GLP-1. I cannot adequately explain the mechanism of GLP-1 to the extent that it deserves, but its important effects are that it (a) stimulates the release of insulin (which is why Ozempic is a drug for type 2 diabetes), and (b) increases feelings of satiety by slowing the rate of gastric emptying.
The latter suggests something profound — people who take GLP-1 receptor agonists will see their set weights change, since they no longer feel hungry after a certain calorie intake. If saying the drugs reduce “calories in” relative to “calories out” helps you sleep at night, it wouldn’t be “wrong,” but the mechanism of GLP-1 receptor agonists effectively changing the set weight explains their success. These aren’t drugs that upregulate metabolism (like fenfluramine/phentermine) — they change how your body desires food.
Incretin-mimicking weight loss drugs will finally provide, to many millions of people, a realistic escape from their body’s high set weight. They don’t require painstaking calorie counting, they don’t require excessive restrictions on food (or classes of food), and they are scalable. They represent a paradigm shift in understanding and resolving obesity as a public health crisis. For that, they cannot be applauded enough.
This is why I find the idea that “diets work, if you stick to them” particularly risible. The fundamental problem of dieting is adherence — if you define “success” as those who adhere to the diet, then you don’t actually examine the factors that make long-term diet adherence difficult for the vast majority of people.
A lot of people who maintain a certain weight without conscious effort often have a pretty substantial day-to-day calorie intake difference. I can range plus-or-minus 500 depending on what kinds of snacks I ate on any given day, or if I didn’t feel like eating breakfast/dinner. This suggests that calorie intake and total energy expenditure are very tightly linked — 1:1, even.