My Experiment with the Personalized Diet

TL/DR; 

I tested my blood glucose levels every 30 minutes for 2 hours after eating. My personal results:  

• Bad: grains (wheat, corn, rice, oats) and desserts cause my blood glucose to spike.
• Good: animal protein (meat and eggs), vegetables, fruit, potatoes, seeds, nuts are best for my blood sugar.    
• Walking for 15-20 minutes after eating any food restores blood glucose to normal levels
• I lost 2" of belly fat in 2 month after doing this experiment, mostly from walking after every meal that I know causes a blood sugar spike (probably small diet change as well). 
• There are other factors than just food that affect blood glucose levels (stress?)

I've been unsuccessfully trying to lower my blood sugar ever since I was told that I have prediabetes.  Even if you don't have prediabetes, elevated blood sugar after eating may cause weight gain and contribute to a host of metabolic diseases.

I came across the research of Dr. Eran Segal and Dr. Eran Elinav from the Weissman Institute in Israel regarding blood glucose response after eating.  The surprising results of their study is that the perfect diet is different for each of us, which makes perfect sense to me.

In this post I'll share the results of an experiment on myself to try and determine my personalized diet.  The approach is to take blood glucose measurements every 30 minutes from when you take your first bite of food.  I used the Care Touch glucose monitor as it had the cheapest cost combined for monitor and test strips (I've taken hundreds of measurements).  All of the plots below include an average measurement over 2 hours (dark line) with maximum and minimum measurements included as shaded regions.  The first plot is for all measurements and the following plots have an overlay of specific meal categories (including or excluding macronutrients or food groups).

I'll describe the first plot in detail to help get you oriented.  The vertical axis is blood glucose level.  The bottom axis is "Minutes after taking the first bite".  Before I started eating, I would take a measurement, that is represented as "0 minutes".  The average reading is a dark line.  The minimum and maximum levels create a shaded region.  All measurements fall somewhere in the shaded region.  All other plots overlay a similar plot on this first plot.  I'll put the plots in a progression of how I have tried to manage my blood sugar levels. Every plot has a dashed gray line at 100 mg/dL, since this is an expected level between meals.  The dashed red line at 140 mg/dL is a recommend upper limit.  You can see that I regularly exceed this limit.

The next plot is of desserts (ice cream, donuts, toaster strudels).  Occasionally I go on a "sweet fast", which is a logical thing to do looking at the higher average and overall levels. By eliminating these treats, I expected it would help lower my levels.

The next plot shows the result of excluding sweets.  This includes levels for all meals that didn't include sugar.  Surprisingly it turns out there are plenty of other meals that still cause a rise in levels.

 Before we figure out what is bad, let's find out what is good, or what foods do not cause a rise in blood glucose above the dashed red line.  My body does very well with nuts, seeds, salads and animal (e.g. eggs and bacon)

The salads all included dressing (mostly ranch dressing).  They also include romaine lettuce, spinach,  mixed greens,  tomatoes,  broccoli,  brussel sprouts,  bell peppers,  red onion,  chicken,  eggs,  carrots,  cucumber, artichoke hearts.  The salad that gave the maximum values also had sunflower seeds,  pepitas (pumpkin seeds), corn, peas and croutons.  The effect of seeds can be seen in the previous plot.

The "animal" meals were meals of only food from animals (e.g. bacon and eggs).  As expected these high protein, no carb meals had no impact on blood sugar.  It's important to note that blood glucose level is only one measure of health.  I'm sure that eating only bacon and eggs wouldn't be considered healthy.

 Meals with protein from animals and any assortment of plants (vegetables, fruit, nuts, seeds, potatoes) were good.  I've known that for a while since I always feel great after these meals.  The exception (excluded here) was pot roast with gravy.

 The following were a couple of unexpected results.  Fruit cause a rise in levels the first 30 minutes and then a drop below the initial levels after 60 to 90 minutes.  For this reason, I put fruit in the good category.

Regardless of what I ate, I made a valuable discovery.  GO FOR A WALK AFTER EATING.  If I exercised lightly after eating (take a walk for 15 or more minutes), I could keep my levels below 140 and my levels would even drop to normal.  Below is a plot for foods for which I measured my levels without exercising and then with exercising.  These include: pot roast, Mexican food, oatmeal, donuts, vegetable lasagna, s'mores.  Note that the time of exercise varied

One of the recommended foods for Type II diabetes is oatmeal.  I try to be healthy with my oatmeal so I eat it with raisins and almonds (no milk, sugar, or other sweetener).   The drop at 90-120 minutes is good but the initial spike was unexpected.  I ended up testing each ingredient separately.  Oatmeal was the prime culprit, raising my levels by around 30 for an hour (I ran out of strips).  The almonds reduced my levels over 90 minutes.  The raisins raised levels only slightly and only at 30 minutes.  
Here are some other results.

Mexican food is worse than average.  An exception was Chipotle's salad bowl (chicken, veggies, brown rice, pico de gaio, sour cream), though I only had one data point (a level of 111 after 30 minutes).

In an attempt to minimize dependencies, I had my wife make whole wheat bread.  She ground kernels of wheat into flour and made bread with water, salt, oil and yeast.  I then ate one slice of bread first thing in the morning with nothing else.  I was surprised at how much my blood glucose levels varied at 30 minutes (112 to 167).  My conclusion is that there is another factor affecting blood glucose besides "what you eat".  For example, the stress hormone cortisol elevates blood sugar and possibly that is the cause of this variation.