Weighty matters

[pozorvlak]

I've put on about a stone and a half in the last couple of years (that's about 20lb to the USians, and 10kg to the SI-users), and it's having all the usual negative effects - half my trousers don't fit any more, there are whole classes of stretches that I can't do any more because the fat gets in the way, and, worst of all, my mother's started commenting on the size of my tummy every time I go and visit. So I decided it was time to lose some weight now before it gets any worse - we tend to run to fat in my family.

Partly on a recommendation from someone (johnckirk?) but mostly because I liked the name, I read John "AutoCAD" Walker's free book The Hacker's Diet ("How to lose weight and hair through stress and poor nutrition!"). I've found it interesting and entertainingly-written, though much of it's undoubtedly old hat to a lot of you. Anyway, here's a summary:

• Hunger is part of a feedback system to keep you about the right weight: when you haven't eaten enough, you get hungry, and when you've eaten enough, you don't get hungry until you need to eat again. Unfortunately, if you're overweight it suggests that your feedback system is subtly misconfigured. But that's OK: if someone has subtly misconfigured eyes, we don't make a big deal of it, we just fit them with some glasses. And engineering has plenty of tried-and-tested ways of fixing misbehaving feedback systems.
• For the purposes of losing weight, we can think of the body as a very simple feedback system: a rubber bag which takes in food and liquid, burns some of them, excretes the waste, and either lays down or burns off fat cells to make up the difference between what it burns and what it consumes (to a limited extent, it will adjust metabolic rate to compensate before laying down or burning off fat - guess what, this bit's probably misconfigured too).
• Hence, if the energy content of what you eat is greater than the amount of energy you burn, you will gain weight at a rate proportional to the difference between them. Conversely, if you burn more than you eat, you will lose weight, again at a rate proportional to the difference. After that, it's just a matter of arithmetic. How much do you burn? How much do you eat? What's the difference between them?
• There are about 3500 calories contained in a pound of fat. So taking my case as an example, I've been over-eating at a rate of (3500*20)/(2*365) = 95 calories per day. That's nothing. That's a slice of toast per day, or 25 fewer grams of rice with my curry, or three pints less per week. Or, looked at another way, a ten-minute run per day. So once I lose the weight, maintaining it shouldn't involve too much disruption.
• However, exercise probably won't be enough on its own - it takes a lot of time to burn much energy, and you'll probably just eat more to compensate. Easier to eat less. Edit: exercise also raises your metabolic rate. Even with this effect taken into account, you still have to do a lot of exercise to have the same effect as even a moderate diet. See my response to shuripentu below.
• While exercise won't be much help losing weight, it's still a good idea - he describes an exercise programme based on the RCAF's 5 Basic Exercises programme (5BX) which only takes about 15 minutes per day and requires no equipment.
• Beyond the raw calorific content, it doesn't really matter what you eat - you can live exclusively on Big Macs and crisps and still lose weight, provided you eat few enough of them. For various other health reasons, you probably won't want to do that. Try to ensure you eat a reasonably balanced diet, but otherwise don't worry too much about it. Anyway, in your quest to stop feeling hungry, you'll probably naturally gravitate towards foods with high bulk and low calorific content - vegetables, in other words.
• Sorted. We now know how to lose weight - ensure that you eat less than you burn. The trouble is, it's quite hard to work out how much energy you burn. We can work out, using published tables and some arithmetic, how much energy is in the food that you are eating, and we can measure your weight, but burn rate is harder.
• Here comes the clever bit: you turn it around, and work out burn rate from your rate of weight loss! If you're eating E calories per day, and losing weight at a rate of L pounds per week, then you're burning E-(3500*L/7) calories per day. Use this formula to adjust E for whatever rate of weight loss you desire - about a pound a week (B - E = 500) is good.
• The trouble with this is that you don't just take in food, you also take in a lot of water, and your change in dry weight from one day to the next will be dwarfed by hour-to-hour variations in the amount of water in the rubber bag, even if you make efforts to (for instance) weigh yourself at the same time every day, after going to the toilet. Your time series of daily weigh-ins is, in other words, noisy data.
• But that's OK, because we know lots of ways of extracting useful information from noisy data. One of the easiest is by taking moving averages: take your last twenty days' weights, add them up, and divide by twenty. Slightly simpler to calculate is an exponentially weighted moving average. Calculate one of these, either by hand or by using a spreadsheet, and ignore the raw data. This both helps you extract the trend, and helps psychologically: no more thoughts of "Oh no! One biscuit and I've put on four pounds overnight! Why, God, why?"
• So, we can work out how much to eat: the only thing that remains is to actually eat that much, and this is largely a matter of willpower (in other words, I'm screwed). But there are things you can do to help: for instance, it's apparently much easier to plan meals in advance (200 calories for breakfast, 400 for lunch, 200 for mid-afternoon snack, 600 for dinner, 100 throughout the day in the form of milk in hot drinks) for instance. Work out some meals that have that many calories in, and then mix and match them. One of the easiest ways of doing this is, surprisingly, frozen microwaveable meals, which have the total calorific content printed right on the box - but you'll probably want to supplement them with some fresh veg or salad or something for fibre.
• That's it, you're ready to start. Get a pair of scales and a logbook for recording weights, plan some meals, and go for it. Try to pick a time when you'll have a couple of weeks uninterrupted to get used to it. And try to pick a time when you'll be quite stressed, so you've got something to distract from the hunger.
• You will be really quite hungry and will generally feel a bit dreadful for the first couple of days, until your body gives up trying to compensate by adjusting your metabolism and starts burning fat. I've certainly noticed this - I felt really quite out of it for the first couple of days. Apparently it gets better soon after.
• Burning fat cells releases various toxic by-products in a process called ketosis: be prepared for bad breath and general stinkiness. Make sure you drink lots of fluids and eat plenty of fibre.
• He also lists some "secret weapons" that will help with cravings - low-calorie but filling food, like caffeine-bearing drinks, bouillon, unbuttered popcorn and dill pickles.
• When you've reached your target weight, don't just stop suddenly: bring it to a halt gradually, reducing your rate of loss over the course of a month or so.
• Once you've lost the weight, you can't just decide "that's it, I don't need to diet any more!". You still have a broken feedback system, so you still need technological help to correct it. If not, you'll just overeat by a few calories a day, not enough to notice your slow weight gain, until suddenly it's a year later and you're a stone overweight and none of your trousers fit again. But the system of weighing yourself and doing trend-line analysis still works - all you need to do is adjust your intake for a rate of loss of 0. You still have to plan meals, but this shouldn't be too much of a hassle by this stage.

It all sounds pretty sensible and straightforward to me. The real test, of course, is whether I'll manage to stick to it, lose the weight, and then keep it off. As Larry Niven reminds us, in one of the corollaries to his seventeenth law, telling friends about your diet won't make you thin, and buying a diet cookbook won't either. But I have the boundless optimism born of total lack of experience.

Apologies in advance to the Two Shaders: I will undoubtedly be rather tiresome on this subject, and hopefully I'll be stinky and halitotic to boot :-)