Friday, January 31, 2020

Interesting Links - 1/2020

Interesting articles I've read over the last month

Diabetes & Quantified Self:

Health & Aging:
  • Fisetin - a nice review of the history and recent clinical trials on Fisetin as a senolytic agent for general improvement of health and blood sugar control in diabetes.

  • Cream Cheese: A Homemade Ice Cream Miracle Worker - Never would have thought of this, but between the slightly sour taste and stabilizers that reduce ice crystal nucleation and growth, cream cheese can be a great substitute for heavy cream or yogurt in ice cream. I tried it out and really enjoy it in ice creams with a Philadelphia base.


Thursday, January 30, 2020

Recipe #3 - Keto Snickerdoodle Custard Ice Cream

Continuing my series of low-carb ice cream recipes (first two here and here), here's one for Snickerdoodle with a custard base. I can't find a record of what I adapted this from, so if anyone recognizes it, please let me know and I'll add an attribution.

This one's a custard base, so it's a bit more work than the Chai Tea (Philly base). That said, it's only an extra 5-10 min. and the creaminess off the egg yolk is totally worth it if that's what you're in the mood for. 

In the recipe below, I give two different options for the stabilizers. If you use the ice cream stabilizer mix, it will stay scoopable even when frozen. If you use the alcohol + xanthan gum, it will freeze solid, but return to an ice cream texture if allowed to warm for 10 min. or so before eating.

Hope you enjoy it,

- QD

Keto Snickerdoodle Frozen Custard

Yield: 6 (3/4 cup)
prep time: 5 Mcook time: 40 Mtotal time: 45 M
Snickerdoodle ice cream using a custard base


  • 170 g allulose
  • 1 tsp. cinnamon
  • 1 tsp. salt
  • 3 g ice cream stabilizer (can substitute 0.25 tsp xanthan gum + 30 g vodka)
  • 360 g heavy cream
  • 355 g almond milk
  • 4 egg yolks (~80 g)
  • 1 tsp. vanilla


How to cook Keto Snickerdoodle Frozen Custard

  1. Set up an ice bath in a large bowl and set aside.
  2. In a saucepan, whisk together allulose, cinnamon, salt, and stabilizer. If replacing the stabilizer with vodka + xanthan gum, don't add them here. Instead add with the vanilla. 
  3. Add cream and almond milk and heat over medium to medium-low heat until the temperature is 170 °F, stirring frequently to prevent scorching.
  4. Using the hot mixture, temper the egg yolks, then return the tempered mixture back to the saucepan and continue heating until ~175 °F and the mixture is thick enough to coat the back of a spoon. 
  5. Strain the thickened mixture through a fine mesh sieve and into a bowl set in an ice bath. This removes lumps and rapidly cools down the mixture to stop further cooking of the eggs.
  6. Refrigerate for at least 4 hours and preferably overnight, then freeze in an ice cream maker according to its instructions


To reduce the calorie content, you can replace part of the heavy cream with additional almond milk (1:1 by weight). I've gone as low as 160 g heavy cream and it's still good, though less rich. Even lower might be possible, but I haven't tried.
Fat (grams)
Carbs (grams)
Fiber (grams)
Net carbs
Protein (grams)
Calculated by adding up macros of the individual ingredients. Allulose not included in the Total or Net carbs.

Wednesday, January 29, 2020

Recipe #2 - Quick, Customizable Keto Frozen Yogurt

My last recipe for Keto Chai Tea Ice Cream got requests for more ice cream recipes on Reddit. Since people are interested, I thought I'd put up the rest of my ice cream recipes. They're in various states of development, so with each recipe, I'll note what I plan to do to improve and then update once I've tried.

This next recipe is for frozen yogurt. It's adapted from a recipe in Genius Desserts, with allulose swapped for sugar and vodka added to reduce the freezing point and improve texture.

What I like about this recipe is that it's extremely simple and customizable. You just mix a quart of yogurt with sweetener, your favorite flavorings, and some alcohol, freeze and you're done.

For yogurt, Two Good brand is my favorite for taste and texture, but any low-carb yogurt will work.

Unlike the Keto Chai Tea Ice Cream, this one does freeze solid and you need to let it warm up before eating if you want a good texture. Next time I make it, I'm going to try swapping ice cream stabilizer for vodka to see if I can reduce the freezing point.

- QD

Keto Frozen Yogurt

Yield: 6 (3/4 cup)
prep time: 2 Mcook time: 20 Mtotal time: 22 M
Super-simple frozen yogurt recipe


  • 900 g (1 quart) full-fat yogurt, I use Two Good brand, but pick your favorite. 
  • 30 g vodka
  • 2 tsp. vanilla or any other flavoring you like (l really liked crystallized lemon)
  • 200 g allulose (or other sweetener)
  • 1/4 tsp. salt


How to cook Keto Frozen Yogurt

  1. Mix all ingredients until fully incorporated.
  2. Cover and refrigerate for at least 4 hours. Note: This step is optional with vanilla or if the flavor is already incorporated in the yogurt. It was necessary when I used solid flavorings, like crystallized lemon.
  3. Freeze in ice cream maker per manufacturers instructions.


This recipe is really flexible. You can swap in your favorite full fat yogurt with whatever flavoring you like. Vanilla can be omitted or replaced with other flavors (I particularly liked adding crystalized lemon).
Fat (grams)
Carbs (grams)
Net carbs
Protein (grams)
Calculated by adding up macros of the individual ingredients. Allulose not included in the Total or Net carbs.
Created using The Recipes Generator

Monday, January 27, 2020

Recipe of the Week #1: Chai Tea Ice Cream

In addition to self-tracking, I also like to cook. I thought it would be fun and hopefully of interest to people to post some of my favorite recipes. All are low-carb, often modified from regular recipes I've found.

I've recently gotten into making homemade ice-cream. This first recipe is my favorite so far. I adapted it from the book Hello My Name is Ice Cream: The Art and Science of the Scoop. I used the author's recipe for cream cheese ice cream, combined with her procedure for infusing a Philadelphia base with tea, and swapped allulose for sugar. The cream cheese and stabilizer give it a fantastic texture. It doesn't freeze solid and remains scoopable (one of the biggest challenges in keto ice cream).

I used Chai Tea from Oaktown Spice Shop, which is near my house, but you can use whatever tea you'd like. The ice cream takes on the same flavor as that tea with milk, cream, and allulose added. 

You can also tune the calories up or down by swapping almond milk for heavy cream, with the expected affect on taste and texture.

- QD

Keto Chai Tea Ice Cream

Yield: ~8 3/4 cup servings
prep time: 15 Mcook time: 30 Mtotal time: 45 M
Chai Tea infused ice cream



How to cook Keto Chai Tea Ice Cream

  1. Whisk together allulose and stabilizer in a small bowl. Set aside.
  2. Add almond milk, heavy cream to a saucepan and heat over medium heat until boiling. Whisk occasionally to prevent scorching.
  3. Add allulose mixture, reduce heat to a low simmer, cook for 2 min., then whisk in cream cheese until melted and homogeneous.
  4. Remove pot from the heat, add tea, and steep for at least 30 min.
  5. Strain through a fine mesh sieve into a bowl, cover, and cool in a refrigerator for at least 4 hours, preferably overnight.
  6. Churn in an ice cream maker according to the manufacturers instructions.


Possible substitutions & modifications: - Chai tea can be replaced with whatever tea you like. The ice cream takes on the same flavor as that tea with milk, cream, and allulose added - The ice cream stabilizer can be replaced with xantham gum, glucomannan, or other stabilizers, but the ice cream will not stay as soft in the freezer and texture won't be as smooth. Still tastes good, though. - Ice cream stabilizer can be reduced by up to 50% and still get a good texture. - Cream cheese can be replaced by milk powder, but you might have to increase the amount of stabilizer to compensate. - You can tune the calories by swapping more almond milk for heavy cream, with the expected affect on taste and texture.
Fat (grams)
Carbs (grams)
Net carbs
Protein (grams)
Calculated by adding up macros of the individual ingredients. Allulose not included in the Total or Net carbs.
Created using The Recipes Generator

Saturday, January 25, 2020

Experiment #2: 48 hours of Blood Sugar Tracking

Figure 1. Measured blood glucose concentration vs. time for both days. Reference bands and lines show target range and high/low thresholds.

Table 1. Summary statistics of blood glucose concentration over 48 h. 

I don't have enough data yet to analyze my food & exercise experiments (see next experiments), so for this week I decided to look at how my blood sugar varies over the course of a day.

Normally, I only check my blood sugar at key times (waking, going to bed, before injecting insulin, before eating, and 1 & 2h after eating). To see if there was anything to be learned from what's happening the rest of the time, I decided to spend 2 days checking my blood sugar every 15 minutes.

Since I have Type 2 diabetes, my insurance won't cover a continuous glucose monitor (CGM). Plus, blood glucose meters (BGM) are more accurate, even according to CGM manufacturers. So, for this experiment, I used my Freestyle Freedom Lite and just measured by blood glucose every 15 minutes.

This ended up being way more useful than I expected. There was a lot more going on between the times I usually measure than I realized. Here's a most important things I learned, my new questions, and ideas for next experiments:

Key Learnings:
  • Shorter testing intervals around key times is extremely informative. My normal routine of testing blood sugar before and 1 & 2h post-meals is not sufficient. There's a lot going happening on both shorter and longer time-scales that will be useful for understanding the effects of different interventions and for optimizing medication. 
  • I'm spending far more hypoglycemic than I had realized, most notably when I exercise. 
  • My peaks in blood sugar from breakfast and lunch occur >3h post-meal and persist until my next insulin dose. This is way longer than I expected and indicates that I need to switch to a longer acting insulin or change something about the meals.
  • My blood sugar drops significantly during cardio exercise (MMA, biking), then returns to normal over 30-45 min. I need to find a way to mitigate this to prevent my blood sugar from going dangerously low.

  • What is happening to my blood sugar between waking and breakfast? Any risk of hypoglycemia while driving to work?
  • Is the long-duration peak in blood sugar after breakfast and lunch real & consistent? If so, what causes it and can it either be shortened by modifying the meal or mitigated using a longer-acting insulin?
  • What is the effect of different types of exercise? How can I mitigate or offset the initial drop in blood sugar during cardio without causing high blood sugar after the recovery?
  • What is the effect of dinner, disentangled from exercise?

Next Experiments:
I'm always interested in ideas for new experiments, so please leave a comment if there's something you'd like me to try.
  • Exercise studies: Measure blood glucose at 15 min. intervals for different duration and intensity bike rides and weight lifting. Also try varying time since last insulin dose and food eaten before exercising.
  • Long peak from breakfast & lunch: Measure blood glucose at 15 min. intervals a few more times after breakfast & lunch. Try varying protein and fiber content to see if those are the causes. Try normal-acting insulin to mitigate.
  • Food & medication studies: It's proving difficult to disentangle the numerous effects going on at any given meal by adding or subtracting particular foods (data too noisy, too many measurements required to observe a statistically significant effect). Instead:
    • Skip meals and medication to measure background trends.
    • Consume individual ingredients (glucose, protein, fiber, etc.) or take individual medications to measure their direct effects.
    • Measure combinations of ingredients and medications to measure interaction effects.
    • This will require more experiments, but I think in the end it will require less time & effort to get reliable results.

Now for the details:

To better understand trends in my blood glucose over the course of a day and determine if there are trends or events that I should investigate further.

General. Blood glucose was measured approximately every 15 min. using a FreeStyle Freedom Lite glucose meter and FreeStyle lancets & test strips. No special precautions were taken to clean the lancing site before measurement. To take a sample, the lancing devices was used to pierce the skin at an ~45 deg. angle from the finger. Blood was then squeezed out by running the thumb and pointer finger of the opposite hand from the first knuckle to the lancing site of the finger. Blood was then wicked into a test strip that had been inserted into the meter and the glucose reading was recorded.

Medication and meals were kept as normal, with the exception of an additional 2 g of glucose eaten with breakfast on the first day as part of an experiment on the effect of glucose on my blood sugar.


Results & Discussion

Overall. The full set of data is shown in Figure 1 with summary statistics in Table 1. Major insights:
  • I had far more hypoglycemic events than I expected, 13% overall and 21% in day 2, with a low of 41 mg/dL. This was extremely disturbing, as hypoglycemia can be extremely dangerous. A closer look at the data (see Evening section), shows this to be from riding a stationary bike, which I hadn't noticed before because my sugar returns to normal levels after 30-45 min. 
  • Other than the low episodes in the evening, and factoring out time spent high due to eating extra glucose as an experiment, my control was pretty good. 
  • Clear dips and spikes were visible from eating, exercise, and insulin, many of which occurred at shorter time scales than I normally test.

To really see what's going on, though, let's zoom in different times of the day.


Figure 2. Measured blood glucose concentration vs. time from 4a-12p, with annotations for key events. Reference bands and lines show target range and high/low thresholds.
  • Waking to breakfast: 
    • The first morning, there was a 10 mg/dL rise, followed by an identical drop over the next 45 min. My normal routine of measuring up arising and before dosing for meals would completely miss this. 
    • The second morning, the rise was the same magnitude (10 mg/dL), but slower, so there was time for it to fall back down before my breakfast insulin. Either way, looks like there's some timing variability here. Need to investigate further.
  • Breakfast:
    • The first morning, I ate an extra 2 g glucose as part of my food & exercise experiments (see next experiments) and saw an immediate rise in sugar, reaching peak after about 1.5h. Very interestingly, the peak persisted for >5h, much longer than I usually test for. 
    • The second morning, I saw an 18 mg/dL drop in the first 45 min. after taking insulin, which then stayed steady for the next hour. This is larger than expected given the only difference from Day 1 was not eating the 2 g glucose, but may be within natural variation. 
    • More significant was the effect of a one hour mixed-martial arts (MMA) class. Since my blood sugar was low, I took 4 g glucose (expected to raise my blood sugar by ~20 mg/dL). From the start to end of exercise, I saw an ~25 mg/dL rise. While this could be accounted for by the 4 g glucose, based on Day 1, I would also expect an ~20 mg/dL rise from breakfast. All together, this suggests a possible 20 mg/dL drop from the MMA class (very speculative, needs more study). On top of that, I saw a brief 15 mg/dL drop after the MMA class finished, which then came right back up. Not sure what this means, but it's consistent with other exercise effects (see evening section)
    • Lastly, the breakfast on Day 2 showed the same long duration peak as day 1, with blood glucose peaking 3.6h after breakfast and persisting for 5 h, until I took my insulin for lunch. I use a rapid acting insulin at breakfast. However, since I made the switch to the rapid-acting insulin, I added additional protein and fiber to the meal. Need to investigate whether I should switch to normal-acting insulin.    

Figure 3. Measured blood glucose concentration vs. time from 10a-5p, with annotations for key events. Reference bands and lines show target range and high/low thresholds.
  • The afternoon was a lot less eventful. On both days, I saw a 10-15 mg/dL drop in the 40 min. between taking insulin and eating lunch, followed by a continued drop over the next 30 min., then a, ~15 mg/dL rise. On Day 1, the rise almost exactly offset the drop, while on Day 2, the drop was ~10-15 mg/dL larger and therefore wasn't full offset. This may be due to the morning exercise on Day 2, or some other source of variation. Need to keep an eye on this.
  • Either way, lunch medication plus food seems pretty well calibrated, which is a bit odd since my breakfast and lunch are identical. Hypotheses to test:
    • There's some other factor causing a drop in the afternoon, offsetting the long peak seen in the morning
    • The long peak at breakfast is related to the dawn phenomenon (additional glucose from liver and/or reduced insulin sensitivity)
    • To test these, could try skipping meals and their associated insulin, which should have different effects depending on the cause of the discrepancy.

Figure 4. Measured blood glucose concentration vs. time from 4-11p, with annotations for key events. Reference bands and lines show target range and high/low thresholds.
  • The fact that I both eat dinner and exercise in the evening makes analysis difficult. Need to run experiments where I do one or the other to better disentangle the effects.
  • Dinner:
    • On both days, my blood glucose increase slightly after taking my dinner insulin. This suggests my blood sugar may still have been rising from after lunch, however, this is a very long time for food to be having an effect. Since my breakfast and lunch are high in protein and fiber, I tried to find information on their direct and indirect effects on blood sugar. However, the information I found was both spotty and ambiguous (especially for fiber, for which I found a lot of articles on using it for diabetes prevention, but no well controlled studies of its immediate effect on blood sugar). Need to test this directly.
    • After the immediate drop, exercise prevents any meaningful analysis of the effect of dinner except to say it doesn't appear to significantly raise my blood sugar. 
  • Exercise:
    • I did ~30 min. of weightlifting on Day 1. During that time, I saw a ~12 mg/dL spike, which came back down immediately after. My blood sugar typically increases at the start of exercise and this is consistent and small enough not to worry about.
    • The effect of biking was more significant. I typically to 2x15 min. stints on a stationary bike in the evenings and previously hadn't noticed any major effects on my blood sugar at my normal 1h and 2h post-meal checks. By testing every 15 min., I can see that, even with taking glucose before starting, these bike rides are dropping my blood sugar by 10-20 mg/dL over the course of the ride, after which it comes back up to baseline over 30-45 min. 
    • On both days, even with taking glucose before hand, the bike rides sent me into dangerously hypoglycemic territory. Notably, I didn't notice any symptoms of hypoglycemia, possibly because they were overshadowed by the effect of the exercise itself.
    • This drop during biking is very worrisome and something I completely missed with my normal testing routine. A few takeaways:
      • I need to run more experiments to figure out the exact details and how to mitigate this effect. 
        • Bike further away in time from peak insulin
        • Use a faster acting and/or better timed glucose to offset drop
      • Since blood sugar can drop then rise quickly with physical activity, I can't rely on just testing before and after meals. For any physically significant activity, I need to test before and after until I'm sure I understand it's effect.
Overnight. I saw a 5 mg/dL drop overnight between Day 1 and Day 2. That's within the normal variance of my meter, so I'll check over a larger number of days to see if it's real. Either way, not significant relative to the other effects I observed.

Conclusion & Next Experiments
See summary above.

- QD

Monday, January 20, 2020

My Routine: Diet, Medication, & Excercise

Updated 1/21/20

Since most of my experiments will be about the effect of modifying my diet, medication, or exercise, I thought it'd be useful to document my baseline routine.
  • Diet:
    • Strategy: 
      • Ultra low-carb (<15 g/day) for easier glucose management
      • High protein (120 g/day, 1.6 g/kg bodyweight)
      • Same meal every day for breakfast and lunch (makes medication tuning easier)
      • Consistent meat and vegetable quantities for dinner (couldn't tolerate 3 identical meals/day, too boring) 
    • Breakfast:
      • 1 serving Ketochow with 2 tbsp. butter.
      • 3 oat-fiber protein muffins
      • 406 calories, 44 g protein, 5.7 g net carb.
    • Lunch:
      • 1 serving Ketochow with 2 tbsp. butter.
      • 3 oat-fiber protein muffins
      • 406 calories, 44 g protein, 5.7 g net carb.
    • Dinner:
      • 300 g meat
      • 150 g low-carb vegetable (spinach, okra, broccoli, cauliflower, green beans)
      • Note: this is the meal I "cheat" on most often. I have a few favorite recipes for which I've calibrated insulin doses and will go out to eat once in a while. 
    • Snacks: 
      • I try to limit this as much as possible (1-2 times/wk) and to offset with additional exercise.  
      • Homemade low-carb ice-cream, cookies, biscuits, etc.
      • Pork rinds
      • High cocoa chocolate (85%)
  • Medication:
    • Strategy: Keep blood sugar as steady as possible (80-90 mg/dL) with a combination of fast- and regular-acting insulin plus a small amount of slow-acting basal insulin to reduce strain on my remaining beta cells.
    • Oral: Metformin (Glucophage brand), extended release, 2x1000 mg (upon waking up and before bed)
    • Insulin:
      • All dilutions are 3:1 (25% insulin)
      • Waking up: 0.5 units Tresiba (slow-acting), 1.5 units diluted Novolog (fast-acting, to off-set the dawn phenomenon)
      • Breakfast: 7 units diluted Novolog
      • Lunch: 4 units diluted Novolog
      • Dinner: 8 units diluted Humalog
      • Before bed: 0.5 units Tresiba (long-acting)
      • Adjustments: 1 unit diluted Humalog for each additional gram of carbohydrate or 56 grams of protein above my normal meal.
  • Exercise:
    • Total time: 7 h/wk, but the martial-arts classes are fun.
    • Strength-training: 
      • 2 days/wk, 30 min. bodyweight upper-body and core
      • 2 days/wk, 30 min. dumbbell upper-body and core
      • 4 days/wk 15 min. stationary bike set at high resistance
    • Cardio:
      • 2 days/wk, 1 h mixed-martial arts (high intensity)
      • 2 days/wk, 1 h kung-fu (low intensity)
      • 4 days/wk 15 min. stationary bike set at high resistance (same as under strength-training)
- QD

Sunday, January 19, 2020

Experiment #1: Measurement Reproducibility

If I'm going to study how external stimuli affect blood glucose, the main measurement device I'll be using will be a glucose meter (the one I have is the FreeStyle Freedom Lite). In order to understand the measurements I make, I'm going to need to know the accuracy and the precision of the device. By law, home blood glucose meters must give results that are within 20% of laboratory tests 95% of the time. That would mean that if someone's blood glucose was 100 mg/dL, the meter could report 80-120 mg/dL. If that's the kind of precision I'll be getting, then I should just give up now. Fortunately, they're a lot better than that. 

Although there are a ton of studies on the accuracy and precision of blood glucose meters, I wanted to assess my specific meter under the conditions that I will be using it. To assess the accuracy, I measured the FreeStyle Control solution. To assess the precision, I measured my own blood glucose in sets of 5 measurements at different times over the course of the week (so that I'd have different glucose numbers).

To evaluate the reproducibility of the FreeStyle Freedom Lite glucose meter under the conditions that will be used for subsequent experiments.

General. All measurements were done using a FreeStyle Freedom Lite glucose meter and FreeStyle lancets & test strips. To reflect my normal usage, no special precautions were taken to clean the lancing site before measurement. To take a sample, the lancing devices was used to pierce the skin at an ~45 deg. angle from the finger. Blood was then squeezed out by running the thumb and pointer finger of the opposite hand from the first knuckle to the lancing site of the finger. Blood was then wicked into a test strip that had been inserted into the meter and the glucose reading was recorded.

Control solution test. FreeStyle control solution was purchased from Abbot on Amazon. Drops of solution were removed from the bottle according to the manufacturers instructions and tested using the procedure described above.

Blood tests. Sets of 5 blood samples were measured according to the procedure described above. The sample set was taken from either 5 locations on the same finger or the same location on 5 separate fingers over the course of ~1 min. Times were chosen to obtain a range of blood glucose values. 

Results & Discussion
Control Solution Tests. Unfortunately, commercial control solutions do not specify the exact glucose concentration, but instead only give a range of acceptable values that fall within the FDA's +/-20% accuracy requirement (link, link). The Freestyle control solution kit I purchased came with two solutions, "High" and "Low". My vial of test strips specifies the acceptable ranges of these solutions as 248-372 and 30-60 mg/dL, respectively. Assuming the ranges are centered on the actual value, that gives actual values of 310 mg/dL for "High" and 45 mg/dL for "Low." I took 8 measurements of each control solution. 

Figure 1. Measured glucose concentration vs. sample # for "Low" control solution. The reference line indicates the expected reading based on the average of the manufacturer's acceptable range.

Figure 2. Measured glucose concentration vs. sample # for "High" control solution. The reference line indicates the expected reading based on the average of the manufacturer's acceptable range.

Table 1. Table of results of glucose measurements of control solution. 

Happily, the readings were reasonably close to the expected value: 2 mg/dL (3%) for the "low" solution and 19 mg/dL (6%) for the "high". The standard deviation and 95% confidence interval were more consistent between the two solutions at 3-4 & 6-7 mg/dL, respectively. As such, if a measurement were run repeatedly, I can expect it to fall between +/- 7 of the observed value 19 out of 20 times. 

Blood tests. Next up was testing the accuracy of actual blood measurements. Shown below are the results of 11 sets of 5 measurements each taken over 7 days at different times of the day. 

Figure 2. Standard deviation vs. average glucose for blood measurements. The grey line and R^2 value are for a linear fit to the data.

Table 2. Table of results of blood glucose reproducibility measurements. 

Surprisingly, the precision of the blood glucose measurements was just as good as those of the control solutions. Looking at Figure 2, there's only a weak correlation between standard deviation and average glucose for a set of measurements (R^2 = 0.21). Given that, I was able to calculate a standard deviation for the whole data set (i.e. pooled standard deviation) of 3.0 mg/dL (95% CI = 6.0 mg/dL), nearly identical to that of the control measurements. This suggests that the variation in the measurements is coming from something in the meter, test strips, or external environment, rather than something in the blood. 

A standard deviation of only 3 mg/dL blood glucose measurements should be good enough to measure the effect of different foods and other external stimuli or interventions. While there is a small correlation between the standard deviation and average glucose, it's weak enough that 1) it may not be real and 2) I can ignore it for now.

Next Experiments
I'm always interested in ideas for new experiments, so please leave a comment if there's something you'd like me to try.
  • I will continue to measure the precision of the meter on a less frequent basis (sets of 5 measurements). Purpose is to:
    • Get a better a more accurate measure of the correlation between standard deviation and average glucose
    • See if there's any trend in standard deviation with time, time of day, same vs. different fingers, etc.
  • To ensure I'm not biasing the sample set by only taking measurements when I'm interested in the reading, I will randomize when I do a reproducibility measurement by rolling two ten-sided die and taking a set of 5 measurements when the number is ≤ 2 (~2/week).
  • For next week, I'll start measuring how my blood glucose is affected by food and exercise. 
    • Getting a clean measurement for these is going to be tricky, as at any given time, there are a large number of factors affecting my blood glucose.
    • To deal with this, I will pick standard times of day where external factors are as consistent as possible and then flip a coin to randomly decide whether to eat or exercise a fixed amount and then measure my blood sugar over time.
    • It may take a while to get enough data to get reasonable statistics. If I don't have enough data by next week, I'll  post an exploratory analysis of the initial data.

- QD

P.S. I'm sorry for the less than ideal graph/table arrangement on this post. I'm still getting a hang of blogger and I can't figure out how to get two pictures to display next to one another.

Sunday, January 12, 2020

And so it begins...

Hello world,

I'm 38 years old, I have Type 2 diabetes, and I'm going to try to figure out how it works and better ways of managing it through self-tracking, experimentation, and (hopefully) collaborating with others doing the same thing.

At the age of 29, I started having the following symptoms:
1.    Insomnia - I started sleeping <4 hours per night, but wasn't tired during the day
2.    Weight loss - went from 210 to ~190 lbs. in 1 month without changing my diet
3.    Thirst - started drinking a ton of water. By the end, ~1 cup every 0.5 h
4.    Frequent urination - presumably caused by #3
5.    Blurred vision - had to sit much closer to the screen at a seminar I attended

I went to the doctor and was quickly diagnosed with Type 2 diabetes. My HbA1c was 8.7%, which means that my blood glucose was averaging >232 mg/dL. People without diabetes will have an H1Ac of <5.7% (average glucose <120 mg/dL). 

I went home and read everything I could find on diabetes symptoms, progression, and treatment. I found out that since most people get diabetes after age 50, there's almost no data on long term survival rates or even disease progression for people my age. 

Not wanting to die, go blind, lose my feet, or even have to start injecting with insulin, I decided to treat my diabetes aggressively, with a goal of keeping my blood sugar the same as that of a non-diabetic. I went on Metformin, started obsessively testing my blood glucose, radically cut down on the amount of carbohydrates I ate (down to <20 g per meal), and cut calories until I hit 165 lbs. (BMI = 23.5).

With all that, I was able to get my blood sugar under control. Here's my HbA1c history for the first two years:

  • 3/2011 - 8.7%
  • 6/2011 - 6.0%
  • 3/2012 - 5.3%
  • 8/2012 - 5.3%

So, I got my blood sugar into the "normal" range, but at the cost of having to think about/worry about food constantly. It was worth it, but very unpleasant.

I was able to maintain this approach for 4 years, but in late 2016 I got pneumonia. My blood sugar was high while I was sick (~160 mg/dL), which wasn’t a big deal, but when I recovered it didn’t come down. With the same diet as before, I routinely had a fasting blood sugar of >120 mg/dL and 2h post-meal level of >160 mg/dL.

After struggling to try to get things back under control with diet and exercise, I went to a series of doctors and eventually ended up going on insulin. This helped, but I was still seeing blood sugar higher than “normal” and since I my blood sugar was “under control” by the standard of care, the doctors I was seeing didn’t want to prescribe any further treatment.

As when I was diagnosed, I did my own search and ended up finding a book by Dr. Richard Bernstein that advocated a more aggressive treatment approach involving a strict and consistent low-carb diet and careful, but aggressive use of insulin to keep blood sugar in the range normal for a non-diabetic. I became a patient of Dr. Bernstein’s and by following his approach was finally able to get my blood sugar back under control to my standards.

Here's my HbA1c history during this time:
  • 11/2016 - 6.7% (after recovery from pneumonia)
  • 12/2016 - 6.3%
  • 2/2017 – 6.0%
  • 3/2017 – 6.4%
  • 9/2017 – Started working with Dr. Bernstein.
  • 9/2019 - 5.1% (most recent)

This new approach works, but at the cost of an extremely regimented diet, exercise, and lifestyle. It also frequently falls apart when I have to travel or eat out for work. Lastly, I’m only 38 and I’m worried about how much more difficult this will get if my diabetes progresses any further.  

A few months ago, I started reading about the Quantified Self movement, people who track information about themselves to better understand and improve their mental or physical well-being. I was particularly inspired by the people who used self-tracking to understand and control diabetes (linklinklink). Thinking about it, I realized that while I had my blood sugar under control, I didn't really understand what influenced my glucose level beyond the simplisitic "more carbs or protein --> higher glucose" and "fiber and sugar alcohols >> sugar and starches".  There's a lot of "common wisdom," as well as peer reviewed research on the topic, but much of it is contradictory, doesn’t apply to people who already have reasonable glucose control, and/or doesn’t provide clear, actionable, conclusions.

So, I've decided to start experimenting on myself to try to better understand diabetes and how to manage it. In particular, I'd like to understand how different external stimuli (food, sleep, exercise, etc.) affect blood sugar levels. My goal is more than just optimization of my own diabetes management, I'd like to try to understand better how diabetes works and hopefully come up with better ways of managing it for everyone. Doing that will require getting others involved to pool data and brainpower. 

That's where this blog comes in. I'm going to post all my experiments here with detailed protocols, results, and analysis. The hope being that others will read it and be able to conduct their own experiments and analysis. If I'm really lucky, maybe we can create a community of diabetes self-trackers that can work together to better understand the disease. 

To start with, I'm going to post once each week. Next week, I'll start with my first experiments to assess the precision and accuracy of my blood glucose measurements.

- QD