Calorie Counting

Life is complicated. My first post was meant to simply be a slightly expanded biography, a way to put some of my beliefs and behaviours into context in a way that is never possible through a few lines of text. I had no intention of making this an on-going blog. It wasn't that I had nothing else to say, but rather too much. Even staying broadly on the topic of fitness, trying to expound further on who I am is a daunting prospect that would likely prove doomed. I can tell you what I've done, what I've read, the basic mechanics of how I got from A to B. People are drawn to a narrative, we understand change, movement, growth. Trying to describe something in isolation, without comparisons, shared references, taking everything from first principles is a much more challenging proposition. There are entire websites trying to explain principles of building certain lifestyles, with thousands of pages of discussion on points that aren't explicit. Gaining knowledge is a process in itself, it takes time. Sharing knowledge is the same and is incredibly difficult to do impersonally.

So it is with low expectations that I begin the attempt to communicate my current point of view on subjects that I think others could benefit from. As much as this is an exercise in ordering my own thoughts I nonetheless hope that some part of it could help someone somewhere. For many people it may not matter at the end of the day, but I believe that it's important to have a correct understanding of things if you possibly can. So I'm starting with nutrition and, specifically, the Calorie Paradigm. This is likely to be by turn tedious and controversial, and not nearly as pretty as my first post. So let's kick things off with a pretty waterfall to distract you:

Fun fact, this place in South Africa claims to be the inspiration for much of Middle Earth.

So, before we get bogged down in the inevitable semantics, let me try to explain what I mean by the Calorie Paradigm. Readers will likely be familiar with the idea that Calories are something found in food. Some kinds of food have more of them than others, like the octane rating of gas. Food is what makes us go. Depending on how you're built and what your level of performance is, the rate at which you use these Calories - your mileage - varies. When your fuel levels get low, you need to eat more to fill up the tank (stomach) so you can keep moving. When it comes to body weight the maths is simple. If you don't put enough Calories in to the tank for the rate at which you use them your gross weight will drop. And if you add more fuel then you're using? Well, you know why they have those warnings not to pull the trigger again once the auto-stop has kicked in on the pump? That's right, obesity.

This is the first point to make, analogies suck. They're an imperfect way to try to convey a foreign idea with familiar concepts, and if you get carried away they break down. Suffice to say the Calorie Paradigm is often shorthanded as calories in/calories out, eat less move more, calories matter etc. It says the the biological processes governing body composition are dominated by calorie balance. And any alternative approach to diet selection (and by diet I merely mean what someone eats, not specifically eating for losing weight) is successful because at root it obeys the governing principles of calorie balance. That's quite a bold claim, but people have that kind of confidence in it because they think it truly is a physical law. On a par with the constancy of the speed of light (though less immutable as calorie balance is eternal). Let's examine that shall we?

So what is the supposed key to healthy eating? The First Law of Thermodynamics:

"For a closed system, in any arbitrary process of interest that takes it from an initial to a final state of internal thermodynamic equilibrium, the change of internal energy is the same as that for a reference adiabatic work process that links those two states. This is so regardless of the path of the process of interest, and regardless of whether it is an adiabatic or a non-adiabatic process."

In vague terms this is thought of as conservation of energy. So if we assume that Calories are simply a unit of energy, then any change in body mass is due to a difference between the energy in and out of the body, because we've all read Einstein's work we are comfortable with the equivalence of energy and mass. To be honest though we don't need go further than the first clause. A closed system means mass cannot pass the boundaries of the system. Every breath you take violates this - and this is not a trivial point. The heat exchange that occurs in breathing is a critical biological feature. People are not closed systems. Thermodynamic Laws are not applicable. And a good job too - otherwise our very existence would be a violation of the Second Law.

But surely there must be some truth to it somewhere? I mean, can't we pretend we're a closed system and apply conservation of energy anyway? Well, ok, we can do some approximations and see what happens, so long as we don't forget that we no longer have the Laws of the Universe backing us up. Let's try expressing this equation in a more complete way. We want to say that:

 (energy in to the body) - (energy out of the body) = change of internal energy of the body

Easy enough to figure out what goes in right? We have incident radiation, hot drinks, trips in elevators, food, punches in the face etc. Excellent. Some of these we can deal with easily enough - it's reasonable to assume that over the course of a day you end up at the same elevation at which you started for example. But others, like the weather, the temperature of what you consume, and the food itself are likely to be variable. So let's just ignore everything except food. Which we'll come back to later.

As for the energy out? Well, aside from the sweating, panting, and heat lost from any number of other bodily fluids, we've got running, jumping, lifting things, standing up, sitting down. There's probably a lot of factors to all that, not least climate, but let's approximate it all as a single number. Then just for kicks, let's simplify changes of internal energy as being equivalent to stored fat. This may seem like a slash-and-burn approach to approximating physical equations, but it's done with a very good reason. It's hard work to try and work out any of those other variables, so the only practical approach is to ignore them and hope they go away. It's not as if the original equation applies anyway (remember? you promised you wouldn't forget).

So let's clear some more of this deadwood out the way and recognise that food is not a unit of energy. It's all the same though right? E=mc² as we brushed against earlier. Well that's not going to get us anywhere. We're not actually capable of converting a sandwich into pure energy. So let's instead use the energy released when you burn the food, and let's use marshmallows instead of sandwiches for our mental image because they genuinely burn. Hell, this is mental enough already, let's have at reality:

 Ok, so we'll convert what we eat into Calories. We don't actually have a belly full of charcoal but it's an approximation right? To be honest, we're too far from the original equations to call this a simplification any more. I'd more properly call this an engineering model. The difference being that engineering models don't claim to be derived from real physical laws. They have their own set of rules which aren't physical, but which through experimentation have been found to give similar results to reality. An analogy if you will. I'm hoping that I've spent enough time now that when I say:

Calories in - Calories out = Change in bodyweight

This is not actually the First Law of Thermodynamics, it's something that looks similar providing we make a lot of assumptions and ignore a lot of other variables. It does not have to be obeyed, and it's only useful in as far as it helps us make predictions about what happens in reality. In fact, a more accurate approximation would have been to stick with conservation of mass:

Atoms in - Atoms out = Change in bodyweight

Here we can just talk about mass, we can justifiably ignore relativity effects and have something more meaningful but less practical. So let's see how practical the Calorie model is.

I should just note here that we made the assumption that the caloric change in bodyweight is represented by stored fat. Of course, there are many instances of people trying to apply this equation to muscle gain. I don't really know why, the maths doesn't even remotely begin to stack up. I'm going to try and keep it simple for now. Which means the first thing we want to calculate is "Calories In". We've already reduced our definition here by specifying that we're counting in Calories. Now let's idealise the situation further and look at this as if willpower were not an issue and people have and are able to exercise full control over what they eat. It's a big assumption, but it could be true for some people so let's give it a chance. We'll even disregard the coupled nature of the equation (the fact that calories expended AND changes in bodyweight are both factors in determining Calories eaten in the typical individual). We should be able to nail down this number right?

Well, if we make a few more assumptions maybe we can. First up, let's assume the packaging is accurate. This is a poor assumption, some foods can be 100% out on their advertised Calorie contents, but we can't really do the maths without it. So take the numbers on the box at face value. Of course, these numbers aren't actually the heat released when the food is burned. They reduce the food to its constituent molecules (approximately) and use standardised numbers you're probably familiar with. 4 Calories per gram of protein or carbohydrate, 9 Calories per gram of fat. These aren't the numbers you get from burning these molecules either, but this time it's a good thing, because they've been adjusted to account for digestion and metabolism! So these numbers are actually an estimate of the energy available to the body from these molecules when they are eaten. Except of course once you combine these molecules with each other, and a bunch of other stuff, and carry out various processing operations, and hell even if you just cook your food before eating it, you end up with differences in how the food is digested and metabolised. And that's just from the food's point of view. These numbers are trying to account for the calories lost directly at the other end of the gut, those 'stolen' by gut bacteria, yet more that is digested but not metabolised and is lost through urine. And if you think these variables don't change from person to person and indeed from day to day depending on health and environment etc. then I'd suggest that's an optimistic view. So those numbers on the box (even if they've been calculated correctly) are an entertaining guess of what the average amount of energy available from a food would be if it were metabolised as such. But there are definitely margins of error. Big honking ones.

So even the most conscientious acolyte would struggle to get a very accurate count of how many Calories they were actually getting in. And as just alluded to, even then this is assuming that all the Calorie-containing molecules are used for energy or stored as such. Of course this isn't true either, the body has many uses for proteins and fats which mean they're not metabolised for energy and so shouldn't really be counted in this equation. I wouldn't blame you for ignoring that either though, it's not as if we have been rigorous up til this point. Let's be exceedingly generous and say there's a 10% error in our intake (though 50%-200% would be quite easy to argue in some cases). Can we do better with working out what our "Calories out" is?

Thanks to the diet industry, we all have a good idea of this one right? Basal metabolic rate (the energy required to stay alive) + thermogenesis (extra energy released as you go about your day actually doing things) + exercise (ok, it's still thermogenesis, but it can be useful to separate out deliberate activity). BMR is the easy one, it doesn't change much day to day. It's also by far the biggest part of where your Calories are spent. As you might expect it's dependent on factors such as body temperature, body composition, size, diet, thyroid function, age, genetics, physical development, drug use and so on. Sound like time for some simplifying assumptions again? Well first let's note that diet cropped up again - Calories in is going to affect Calories out. You can't simply change one without changing the other. But let's ignore that. It turns out that there is remarkable consistency in metabolic rates. Most of us our within 10% of each other, even without doing any maths. That's handy. Not terribly precise, but if we correct for mass and age at least we can probably get a bit closer, probably within 200 Calories.

This is a good time for a side note. An extremely common suggestion for people wanting to lose weight is to gain muscle, as more muscle means higher BMR. I'm not saying it's not true. I'm just saying that if you spend a year losing 25lb of fat (quite feasible) and gaining 25lb of muscle (quite challenging) then you've boosted your metabolism around about a whopping 100 Calories. 

So, have much variation is there in thermogenesis? Well, depending on your level of activity the calculators will suggest anything from 10% to 100% increase on your BMR. Even if we all burned Calories as efficiently as each other, in all our varied definitions of activity, this is still highly dependent on self-reporting. And this is where the diet factor will really kick in. If you have energy to burn, you're likely going to burn it. If you don't have as much available, you can sweat at the gym all you like, but when you get home you're more likely to sink into the couch as your BMR ratchets down to keep things on an even keel. It think it's doing you a favour. This again actually works for us, it reduces the variation between people so it's easier to guess what the total energy cost is. Because it's surprisingly consistent within and between cultures. Even so, believing that you know better than within 10% accuracy what your daily energy expenditure is would be a bold statement. The estimates are good because statistically there's not much variation, not because the calculators really mean anything.

Let's try and use our equation now to lose bodyfat. Everyone knows a pound of fat contains 3500 calories. Sure, why not. So if we want to lose a pound of fat, we just need to plug in our "Calories out" and it's easy to then see how much we need to try and eat. Except it doesn't work like that. Barely anyone will try and persuade you it's that simple any more. You can only lose so much at a time. Why? The Calorie model doesn't say, we're just adding an additional caveat, changing the rules, which is fine since we made the rules up in the first place - it's our model remember? Not the Universe's. Empirically then we can say losing a pound of fat takes about a week. So we want to carefully balance out a 500 Calorie deficit. We might as well just make that the rule. If you want to lose bodyfat, calculate a 500 Calorie deficit. It's not as if we expect the model to hold up for weight gain at this point, not without additional empirically-derived caveats. Let's suppose a typical metabolism of 2000 Calories. It's on the low side, but we should probably include kids in our averages these days right? Now taking our optimistic tools, we think we can achieve a deficit of 150-850 calories. We can't really know any more accurately than that. At least we've guaranteed a deficit with our model, though you may be disappointed to be stuck at a third of the rate of progress you'd hoped. At this point it would probably be best to monitor your bodyfat, and if you're not losing enough then try and eat less, and if it's dropping too fast eat more.

I'll be honest, my atom balance equation isn't looking such a bad idea any more. Give me half an hour, or decades of research and experiments, I could probably suggest a mass-based intake that would get you in the ballpark and then have it self-correct by adding or reducing the amount of mass you consumed depending on the results. And without all this tedious framing of everything in Calories. Not that I'd recommend it as a strategy, because it still fails to address most of the issues that cause people to fail to adhere to a diet in the first place. I'm just saying that the whole Calorie paradigm is no more meaningful than trying to eat a specific weight of food a day. We could use any other property of food as well with decent effect. Ok, not all grams are the same, we should probably specify that it will work far better if you eat certain kinds of foods. Eating your allotment in paper every day is not going to work well, whether you measure it kilograms, Calories or candelas.

(thanks google, that's exactly what I was looking for...)

So if I think the model is fairly flawed (and I do by the way), why does it seem to work? Well, it works sometimes, but then it would because some people who are trying to lose weight will eat the right things and do the right things. And in fairness the underlying reality may have a strong correlation to mass of food consumed (and if you try to reduce calories, you will likely reduce total mass eaten - you'll also spend less time eating, and eat less of all macronutrients). It's not too much of a stretch to say that obesity may be linked to overconsumption in some aspect or other, so anything that reduces intake across the board has a chance of working. But that doesn't mean there's anything special about Calories. And before we get sidetracked again into Calories representing a fundamental energy balance, may I refer you to the bulk of this post above?

More significantly, simply the act of tracking something would have improved the diet. You can assume any explanation you like here, whether it's being more mindful of your eating, making better choices, becoming more educated about what's actually in your food. Tracking has been seen to work with nothing more advanced than plotting your weight daily against an arbitrarily determined slope. Without obsessing over any behaviours at all. So if you start digging up information and writing down every constituent of everything you put in your mouth, you'll probably lose weight if you want to lose weight, and gain if you want to gain. Unless you're trying to hit a certain number of course - and that number ends up being wrong. So it's probably lower risk to skip the calculators altogether.

There's a bunch of additional caveats these days to try and make the model work of course, some of which address the problems of adherence (hunger etc.) while others try to cover for the errors inherent in the calculations. Some do both. As an example, the Calorie model will often now include a rider to the effect of increasing protein consumption. Often to twice the amount or more than would appear to be required by any scientific study. This wouldn't be included unless it helped right? Or if there was a lot of money at stake in supplementing protein I guess, but what are the odds of that? So protein is quite satiating. Getting protein Calories will make you less hungry than cupcake Calories. What happens if we forgo the Calorie Paradigm and try phrasing that differently. Protein will make you less hungry than cupcakes. Hmm, seems to work, and we can probably explain it with some kind of biological explanation that has nothing to do with thermodynamics. Cool. Is excess protein a handy recommendation in other ways? Well, yes. As noted previously, protein isn't really a fuel at all, most of those Calories ordinarily wouldn't count to any energy balance. What happens if we double the amount of protein? Well, if you're desperately short of energy it can be metabolised, but if not it can just be broken down and the nitrogen excreted along with the rest of the waste. It's insurance if you like. You're eating Calories, but the amount you actually use depends on whether you need them or not. It's almost as if your body doesn't need you to get out your lab equipment and micromanage your meals. So long as you follow some simple rules it can balance itself out. Those rules are probably a subject for another time, but they probably won't need to justify everything through the use of Calories. In fact we could probably avoid a whole load of question and misunderstandings and stress if we didn't mention them at all. Just a hunch.