I take a data-driven approach to my health and find that it's a pretty straightforward task. I write an exhaustive post of all the things I've read/done in order to create a resource designed to demystify health, diet, and fitness. The goal is to demonstrate that managing your health is surprisingly quantifiable, predictable, and easy.

This document is very long (>5000 words), and a lot of it is not necessary to read. This is more like a big folder of my health notes, the basis for my decisions, my progress, and a rough tutorial. I've tried to structure it in a manner that makes it easier to read depending on your goals. If you're interested in how I approached weight management, you can read from "Intro" to "Some Concerns". If you're interested in how I approached getting fitter, you can read from "Aerobic Exercise and Resistance Training" until the end. If you're just interested in a data-driven, evidence-based approach to health, you can read the whole thing.

During my last year at University, I gained weight. I had recently come off of essentially solving a long-time sleeping problem (hooray for blood tests!) and wanted to take an extremely hard course schedule to make up for lost time. Swamped with work, I ended up skipping exercise and eating much more frequently, to the point where I ended up slowly making my way up from 145 lbs to 185 lbs.

Fast forward to a few weeks out from graduation, and I've become the heaviest I've ever been. Resolved to solve this thing I considered a problem, I did some reading (with google scholar searches). It's particularly difficult to find good science on human nutrition, since most of it seems to deal with fad diets and new products rather than simply finding the optimal path to weight loss.

Some papers that did interest me were the following:

The Role of Exercise and Physical Activity in Weight Loss and Maintenance (Swift et al. 2014)

• Unless volume of aerobic exercise is high, clinically significant weight loss is unlikely to occur through just exercise.
• Exercise plays an important role in preventing weight gain by raising energy expenditure
• Extremely dramatic results in several groups over a year
  • Pedometer-based steps goal: 0-1kg lost
  • Aerobic exercise: 0-2kg lost
  • Resistance training: 0kg lost
  • Aerobic and resistance training: 0-2kg lost
  • Caloric Restriction and aerobic exercise training: 9-13kg lost (20-30lbs) (!)

Micronutrient quality of weight-loss diets that focus on macronutrients: results from the A TO Z study (Gardner et al. 2010)

• Studied Atkins, Zone, LEARN, and Ornish diet groups, all women
• After 8 weeks, a significant number in a few of the groups had shifted to deficient levels for some micronutrients

Is there an optimal macronutrient mix for weight loss and weight maintenance? (Wilkinson et al. 2004)

• Low-Carb diets were found to enable short-term weight loss, since they facilitated reduced energy intake, but the lack of long-term studies on them make them difficult to recommend
• "Successful weight loss occurs through the creation of a sustained energy deficit, and should be achieved through a combination of exercise and a nutritionally balanced and varied diet."

Replacing Sweetened Caloric Beverages with Drinking Water Is Associated with Lower Energy Intake (Stookey et al. 2007)

• "The caloric deficit attributable to replacing SCBs with water was not negated by compensatory increases in other food or beverages" which is pretty straightforward
• "Replacing all SCBs with drinking water was associated with a predicted mean decrease in total energy of 200 kcal/d over 12 months."

Water consumption increases weight loss during a hypocaloric diet intervention in middle-aged and older adults. (Dennis et al. 2010)

• Drinking water before meals generally leads to less food intake compared to not increasing water intake
• "Drinking water reduces sensations of hunger and increases satiety"
• (This study was done with self-reporting so I don't assign the strongest confidence with it, but the effect was a noticable +44% so I included it here)

Variability in energy expenditure and its components. (Donahoo et al. 2004)

• Variation in energy expenditure is 5-8% for BMR, 1-2% in exercise expenditure, and 20% from diet-induced thermogenesis, all very reproducible
• That is, 68% within 1840-2160kcal/day, 96% within 1680-2320kcal
  • Less than 1% chance of comparing two people and finding them to have a 600kcal or higher difference in BMR; most people differ by roughly a single poptart.
• Ergo, variation in humans is largely due to variation in non-exercise activity (e.g. occupation, average time spent walking, whatever)

And then some of the more classic studies:

Calorie Intake In Relation to Body-weight Changes In The Obese (Kekwick & Pawan, 1956)

• Commonly cited paper used to argue for low-carb diets
• Obese individuals were placed on different macronutrient heavy diets: one 90% carbohydrate, one 90% protein, one 90% fat, each at 1000 calories
• Protein and Fat groups lost weight, carb group seemed to actually gain weight
• Wasn't actually three groups on different diets, was actually the same individuals doing each of these diets for around 1 week, with weight loss stalling during the carb-heavy week
• "The rate of… loss of water has been shown to rise with high-fat and high-protein diets and to fall with high-carbohydrate diets."
• "The rate of weight-loss has been shown to be proportional to the deficiency in calorie intake when the proportions of fat, carbohydrate, and protein in the diet are kept constant at each level of calorie restriction."
• Possible sources of error included patients having a tendency to lie and eat extra food (referred to as "inadequate personalities"), as well as women whose water retention varies with the menstural cycle
• Kind of flimsy experimental design if you ask me

The Biology of Human Starvation (Keys et al 1950)

• Famous study where participants were starved for 24 weeks, only eating relatively small amounts of carb-heavy foods to simulate wartime starvation (potatoes, bread)
• Participants also walked 22 miles a week and did physical labor work
• Weight loss was very stable with a consistent caloric deficit and would plateau after 24 weeks, whereupon the participants looked like actual skeletons
• A bunch of horrible psychological and physiological effects on the participants was found after starving them for 24 weeks
• Not super sure why this is so frequently cited with regards to healthy dieting, aside from the "plateaus will go away if you keep going" implication, and the fact that slowing metabolism happens roughly quadratically, with a true plateau at 24 weeks
• Please don't starve yourselves for 24 weeks

• Fad diets generally seem to work by "tricking" you into eating fewer calories by eliminating calorically-dense foods, with some of the carb-restrictive ones offering an encouraging boost at the beginning that comes with reduced water retention.
• There's nothing "inherently" dangerous about most foods restricted in these diets, merely the amount of them that you eat relative to other foods.
• Fad diets can encourage short-term weight loss but if you "finish" them and go back to normal you'll probably just gain weight again.
• Fad diets can cause nutrient deficiencies, at least if done improperly.
• The "solution" all of these diets seem to be hinting at is just caloric restriction. Intermittent Fasting plays upon this by only letting you eat in short windows (and hopefully less overall), low-carb plays upon this by removing stuff like bread which has lots of calories, etc etc.
• Most people don't have vastly differing metabolisms, overwhelming odds suggesting I am not "predisposed" to losing or gaining weight.
• Exercise doesn't necessarily make you lose weight (!), eating less does.
  • Exercise obviously has other benefits and you should still do it, this is more to say "diet is what is important with regards to weight".
• Drinking only water is a really easy way to get a free ~200kcal/d off your diet, if you drink sweet beverages often.
• Drinking water might make you feel more full, so drinking a lot of water could make the sensation of being hungry much less salient. Good hydration is obviously important for other reasons as well, like short-term memory and visuomotor tracking.
• To lose weight in the most consistent way, you should measure your caloric intake and eat a consistent amount of food every day.

Cronometer is a website that allows you to track your meals, your nutrient intake, your exercise, and your weight. You can export your data to CSVs, you can ask it to send email reminders at set times everyday, and you can adjust pretty much any parameters you want. I started tracking all of my meals on May 26th 2017 (although I had been going through C25K beforehand and had already lost some weight). When I started, I was at around 175 lbs and began with a moderately gentle -500 caloric deficit, although I later dropped it to around -1000. I used mostly the default values for every category, and tried my best to make everything green (i.e. get enough of every macro and micronutrient, every day, while maintaining a caloric deficit and a reasonably consistent macronutrient ratio). I weighed myself every morning as soon as I woke up, at around the same time each day. I started out extremely manic about my intake (measuring food with a scale to ensure I wasn't over or underreporting) but as time went on I grew better at estimating how much food I was eating (mostly by weight and portion control). In general, I ate anything I felt like eating, provided I had enough calories left in the day to eat it. Their iphone app costs $3 (and to be honest kind of sucks, frequently crashes and doesn't load) but you can use the webapp for free and suffer through their mobile site on chrome if you don't want to spend the money.

I actually didn't know there were other biometrics that weren't weight until 6/27/17, which was 32 days into this process (and -11 pounds). From then on I started tracking mood (measured in "milligiggles") and sleep, mostly in hopes of getting good data that I could use in the future. Tracking all these gives me a wide range of variables that I could use for fun projects in the future (Am I happier when my calorie intake is higher? Am I more likely to go running on days where the weather is nice? Do I sleep more when there's low barometric pressure? Am I happier on days where I run? The possibilities are finite but reasonably interesting, and gwern did a similar project where he found that weather had no effect on his mood or productivity). On top of that, I bought an omron scale which I started using on July 5th, 2017. This was half because I felt I ought to own my own scale (I had been using my family's) and because it would allow me to track body fat percentage, skeletal muscle, and visceral fat. It was also pretty fun to mess around with the data points and see their accuracy (Is there correlation between amount of water drank vs bodyfat % measurement? No. How much does my weight fluctuate throughout a normal day? Around 2.5lbs. Is visceral fat or skeletal muscle useful to measure? Remains to be seen.)