Week #19: 297.0 lbs (37.0%)

Christmas was evil as well! Time to redouble my efforts for the new year! 2015 is it!!!

Week #18: 292.6 lbs (37.2%)

This is not good!

Week #17: 291.0 lbs (35.1%)

Well, at least it's going down: lost 2 lbs this week! The blue curve is overall and the red curve will continue to be after the holidays.

Week #16: 293.0 lbs (34.0%)

The holidays are evil! Started regaining at Halloween and Thanksgiving was horrible! Back to the grind stone!!! The red line was my original slope, the blue line is my slope now, so I'm about 30 lbs behind.

Week #10: 279.6 lbs (32.6%)

Not a good week! Went up by 2.4 lbs! Must have been those donuts I ate at Physics Forum! Gotta fix that!!!

Week #9: 277.2 lbs (32.8%)

Plateau, here I come! Only 1.4 lbs this week.

Week #8: 278.6 lbs (33.5%)

Only 2.4 lbs this week. Approaching a plateau? Better keep chugging along!

Week #7: 281.0 lbs (33.5%)

Down 3.6 lbs this week. Hopefully I get back up to around 5 lbs/week next week! 280 is the lowest I've been in the last few years; I should be there next week! I got rid of the other curves as they were meaningless at this point. Still predicting late January or early February for my goal of 220!

Week #6: 284.6 lbs (34.5%)

Lost 6.4 lbs this week! Still on track to hit my goal near the end of January.

Week #5: 291.0 lbs (33.0%)

Only down 1.2 lbs this week, but at least it was down! I spent three days this week on a business trip to Pittsburgh, and boy was is hard to be "good." My next trip isn't until Thanksgiving, so I should be able to make a lot of progress between now and then! My equation is now predicting that I'll be at my goal weight on Feb. 1st.

Week #4: 292.2 lbs (34.8%)

6.4 lbs this week! (But my body fat % went up. Oh, well!) Hopefully this trend will keep up! If it does, I should be at my goal weight by January 5th!!! Here's the graph:

Week #3: 298.6 lbs (34.2%)

Below 300!!! I'm hoping this trend keeps up! Neither of the predictive models is working very well, so I'm creating my own! It makes sense that calories burned per day is proportional to weight (one could argue that it's linear, but we'll go with proportional for now). Under this assumption, the equation is:

dw/dt = 1706 – cw(t).

Solving the equation, and using the data thus far to solve for c, we find that c = 15.4767. This gives the yellow line in the graph below. The equation predicts I'll lose 5.6 lbs this coming week, so we'll see how close that is!

Week #2: 303.2 lbs (37.2%)

Down 6.8 lbs for week 2! I don't think its really valid, though, as I worked on the house yesterday and am pretty dehydrated. But I'll take it anyways!

Neither prediction is doing very well (the blue and red curves). The yellow curve is a fit to the equation w(t) = ae^bt + 130, and I only fit the equation to the last two points, because I think the first measurement is inflated relative to my current diet. If I include the first point and leave the asymptote unknown, Mathematica chokes! I used 130 instead because that is the number from the red curve and it seem reasonable. More data will always help!

Week #1: 310.0 lbs (37.4%)

One week down! I lost 11.4 lbs this week and 2.5% body fat, but I'm sure more than half of that was water. So much for the Mifflin St Jeor equation! A few years ago I found another equation that said I would burn about 12 kcal per lb, so if I add my walking to that (and use today's weight as the "initial" condition) I get another function that I think is more reasonable, and that I expect to trend well going forward:

w(t) = 130.229 + 184.468e^{–0.0257947t}, where t is in weeks.

If I plot this equation, the Mifflin St Jeor equation, and the data points, I get:

So, according to Mifflin St Jeor, I should lose only 2.0 lbs this week, but the new equation predicts 4.6 lbs, which sounds more realistic to me! I'll keep an eye on these curves, as well as a regression fit, into the next few weeks.

Week #0: 321.4 lbs (39.9%)

It seems that everyone nowadays is blogging about and posting their weight loss progress (or lack thereof) on Facebook and Twitter. I think this is great for accountability purposes! The problem I have always had when attempting to lose weight was the periodic cheating, and that the only one who knew I was cheating was me. I need someone to yell at me when I screw up!

Since I am an engineer by training and at heart, I am taking a very logical and mathematical approach to all of this. My goal weight is 220 lbs. For my height, 6 ft 4 in, that is the upper bound of my ideal weight range, and my actual body fat percentage should be in the normal range at that weight. (Doctors tend to follow the government's advice and use BMI to tell if you're overweight. BMI is only intended as an approximation to body fat percentage, so if you have a scale that can measure body fat, that is a better estimate anyways. My BMI at 220 lbs will be 26.8, which is still considered overweight, but I expect my body fat percentage at 220 lbs will be around 22%, which is considered healthy.)

As I am an engineer, I would like to define a strict regimen for diet and exercise (a.k.a., a procedure!) that I can use to project my progress. This will consist of the following menu each day:

• Breakfast (570 kcal):
• 53 g Kashi GoLean Crunch cereal (200 kcal)
• 1 cup whole milk (150 kcal)
• 1 cup orange juice (110 kcal)
• 1 medium banana (105 kcal)
• 2 cups coffee (5 kcal)
• Lunch (364 kcal):
• 2 slices Freihofer's double fiber bread (200 kcal)
• 1 Tbsp spicy brown mustard (15 kcal)
• 65 g Inspirations deli turkey (77 kcal)
• 1 can V-8 juice (72 kcal)
• Snack (165 kcal):
• 12 fl oz hot tea (4 kcal)
• 1 tsp honey (21 kcal)
• 1 Kashi Peanut Peanut Butter bar (140 kcal)
• Dinner (607 kcal):
• 1 SmartOnes frozen meal (282 kcal)
• 16 fl oz whole milk (300 kcal)
• 12 fl oz herbal tea (4 kcal)
• 1 tsp honey (21 kcal)

My total daily caloric intake will be 1706 kcal, on average.

For the loss term, I will use the Basal Metabolic Rate computed using the Mifflin St Jeor equation: 10w + 6.25h – 5a + s, where w is weight in kg, h is height in cm, a is age in years, and s is +5 for males and –161 for females. This equation gives the number of calories burned per day with no additional exercise. For my specs, this will be (4.55w + 1036.5) kcal/day, where w is weight in lbs.

I will also commit to walking at lunch each work day. On the weekends, Ashley and I are working on the house, which is a different kind of exercise. A moderate walking pace is 3.3 METs, where 1 MET is the equivalent of burning 1 kcal per hour per kg of body weight. That means my daily walk burns 0.8w kcal/day, where w is weight in lbs. This makes my total loss term (5.35w + 1036.5) kcal/day.

Now I have enough to form a differential equation for my weight (production – loss):

dw/dt = 1706 – [5.35w(t) + 1036.5] = 669.5 – 5.35w(t) kcal/day

What I really care about is lbs per day, not calories per day. One pound of human fat tissue contains approximately 3,555 calories, so assuming the weight loss is all fat (which I hope it is!), we can rewrite the equation as:

dw/dt = 1.318 – 0.01053w(t) lbs/week

So, in my first week, I should lose about 2.07 lbs! If we solve this equation for w(t), we find:

w(t) = 125.166 + 196.234e^–0.01053t lbs

This means I should expect to hit 220 lbs in about 69 weeks, or 1 year, 4 months! That's a lot of dieting! I hope my assumptions are wrong!

I will also be trying to stay well hydrated by drinking 1 fl oz of water for every kg of body weight. I am including coffee, milk, tea, and orange juice in this "water," and I will be updating the daily quantity weekly. The first week will then include 74 additional ounces of water each day.