HFCT is the acronym I’m using for a high-frequency cholesterol testing study I’m conducting on myself (along with a group of QS-ers conducting similar studies) in the fall of 2017. This is my 4th blog post about it. Start back at the first to catch up.
Friday I started the third and final experimental segment in my HFCT project. I had a bit of a break in the last week because I was traveling for a conference, so I took the opportunity to reflect a bit on how this HFCT thing has been going so far.
getting started: fruitless good intentions
I had about a week between returning from a two and a half week trip involving four countries in Europe1 and the start of my first dietary protocol in my HFCT study. I had hoped that by the time I was due to begin my HFCT study with a “baseline” day of every-two-hours testing and my very low carb segment of my experiment starting the very next day that I would have planned out my meal options along with shopping lists, have shopped already for the first few days or even a week of the experiment, and already be more-or-less keto-adapted. To achieve this last, I had planned to start eating very low carb as soon as I returned to the States.
In reality, no part of these well-intentioned plans came to fruition. Returning from my trip and jumping back into work after such a long absence occupied more of time that I expected, and so did additional obligatory preparations for the study that I hadn’t thought through: meeting the study coordinator, Azure, to pick up the remainder of my supplies and finalize my experiment schedule, practicing with the CardioChek Plus and scheduling and fulfilling a lab appointment to validate the device against a venous blood draw lab lipid panel.
The adjustment back to the Pacific timezone after my travels in Europe also threw a wrench in my keto-adaptation plan: my insulin needs were quite unpredictable—due to my Circadian rhythms being in timezone-readjustment flux—and I had to resort to treating several episodes of hypoglycemia with glucose, preventing me from achieving ketosis.
So in the end, I kicked off my very low carb segment with a clear liquids-only fast for about thirty hours prior to the start of my “baseline” every-two-hours testing day. This succeeded in rapidly dropping my metabolism into a ketone-dependent state, but I think it contributed to a rather rough start to the experiment, since I effectively went cold turkey on carbs to zero with the fast and then nearly zero while eating very low carb. During the first five days or so when I was eating very low carb, my insulin needs dropped so low I was struggling to keep my blood glucose up.
ongoing: the incredible time-sink of food prep
By far, the biggest struggle I’ve had in completing the first two dietary protocols in my experiment spanning about a month of time2 has been time. I hadn’t realized exactly how much time it would take to be preparing all of my own food and tracking it carefully. Possibly I hadn’t even realized the extent to which it is actually rare for me to prepare food at home (aside from very simple things like eating my fairly typical breakfast of coffee and a 125-gram serving of plain Greek yogurt). At first, I was cursing myself for not having had the discipline to plan out more of my meals and shopping lists in advance, as I had thought that would make it easier. But in hindsight, I’m not even sure that would have helped. I simply don’t have the time in many, many of my days to prepare food before work (for breakfast and lunches) or after work, often because I’m not home in the evening3.
All in all, pretty much everything that could go wrong with my food preparation and meals during my very low carb and low carb protocols did…
On the bright side, however, for my final high-carbohydrate dietary protocol I don’t think I’ll have quite the same degree of difficulty with food prepartion time. Carbohydrates are a lot easier to prepare in advance and store without loss of quality than the kinds of meals I prefer when eating low carb (freshly put together salads or cooked meat or fish).
Sadly, I wasted a bunch of food. Since I hadn’t planned out meals or shopping lists, in my last minute panic and rush to start the project, I made a trip to Trader Joe’s and bought a whole lot of the food in my allowed food groups, without thinking through exactly how and when I was going to use some of the vegetables and other perishables…and some of them were well past their best-by dates and/or had obviously gone bad by the time I got around to them, and I had to throw them out.
keeping types of fats constant
Originally my plan to try to control for different types of fats was to aim for one meal heavy in animal saturated fats (e.g., containing a full serving of bacon, or red meat, or chicken with skin, or pork) and one meal of fish. Just because of how I had to juggle what was in my fridge, what I could easily bring to work4, and what I had time to cook, I had to throw out this idea very quickly. So it’ll be interesting to see how much the types of fats varied across the different days of my protocols and whether that appears to correlate at all with the variation in my lipids. I know there was at least one day when I had fish-centric meals for both lunch and dinner, and the opposite probably happened as well.
eating only what I prepare
I had planned not to eat out at restaurants at all during my experiment, but I caved on that quite quickly, for two reasons. First, I didn’t want to be anti-social. So I ended up having three lunches out with a current coworker and two former coworkers who were in town visiting, and I just resolved to take a picture of each meal and do my best to try to estimate the macronutrients later. Second, I ended up buying lunches out from chains (mostly Mixt salads) near my office that provide nutrition facts for their foods online simply because I couldn’t scrape together the time to shop for enough supplies for my fridge to pack for lunches and to pack those lunches. After trying a Mixt salad once and judging the nutrition facts to be accurate based on how my blood glucose behaved after eating the salad, I resolved to “cheat” in this way as a way to ensure I’d actually be able to complete the experiment without going nuts.
shopping for the high-carbohydrate dietary protocol
I did manage to plan some meal ideas and shopping lists for my final high-carbohydrate dietary protocol, and I’m glad I did. Since my goal is to recreate the type of diet I was prescribed upon being diagnosed with type 1 diabetes and even more specifically, to recreate the typical foods I used to eat prior to my first experiment with a low-carb diet in 2011, there were very specific things I wanted to find, and I needed to organize my shopping across the various grocery sellers I’d need to visit to obtain everything. There were only two surprising setbacks in this process:
- Trader Joe’s didn’t have the dried tart Montmorency cherries that are my favorite part of the steel-cut oatmeal breakfast5 that was one of my staple breakfasts prior to 2011. I’m hoping they come back in stock before this section of my dietary protocol ends, as there’s no way I’m going to keep eating oatmeal for breakfast after this experiment concludes6 but I do want to recreate this breakfast, and the tartness of the Montmorency cherries is a crucial component.
- It’s been so long since I’ve purchased sandwich bread that it took me about ten minutes to find at my local Whole Foods 😂; it’s quite possible I have never purchased any sandwich bread at this Whole Foods, ever, and I moved to California four years ago!
P632 vs. P718 lipid panel strips
At the very end of my second dietary protocol, I opened my third box of PTS lipid panel strips, and this box (and my fifth and final) is lot number P718 instead of P632. To my complete shock and surprise, the first time I did a test with a P718 strip, I got an HDL value over 100 mg/dL, not just “> 100 mg/dL”! This prompted me to pull out the paper insert from the strips, and sure enough, this slightly different strip7 handles an extended HDL range up to 120 mg/dL. My concern now is whether I should use some of my extra strips (see below) at the end of my experiment to do a precision test comparing side-by-side P632 (assuming I can get some help scrounging up some more) and P718 tests, to make sure that all other results like triglycerides are precise between the two strips.
The fact that I didn’t know before now that the P718 strips with the extended HDL range exist is a bit frustrating. Since I come close on most tests and have actually topped out the P632 strips (just getting a non-numeric “> 100” reading for HDL) a handful of times during my experiment, it would have been really valuable to have had five boxes of the P718s!
“real life” tests
I’ve had to break up my experiment due to a couple of planned trips. Originally, when Azure from Quantified Self and I were working out my testing plan, she recommended bringing the CardioChek Plus with me on my trips, and we scheduled a few “real life” morning fasting lipid panel tests during these trips. I did bring the device with me on my first trip to Toronto for a conference, and I managed one of the real life tests because I arrived in Toronto a couple of days early to see the city and thus had greater freedom the first couple of days to time a dinner and following breakfast leaving at least a twelve-hour gap between them. After the conference started, though, this became intractable, and so I skipped the second planned “real life” test, and I suspect the same will be true for my next planned trip (for a family event), and so I don’t think I will even pack the device. (I also feel like there’s limited utility in a test that I have no dietary context around since I’m taking a break from food prep and logging while traveling.)
Germany, Austria, Estonia, and Hungary, for the curious. ↩
My third and final dietary protocol will stretch longer, as it’ll be interrupted by some planned travel for a family event when it won’t be at all feasible to track everything I’m eating to the level of precision I want. ↩
I take ballet classes as a hobby and signed up for an unusually high number of classes this fall—usually three evenings during the work week. I’ve been loving the ballet, but the added busy-ness in my schedule has been very tough to reconcile with this HFCT project. ↩
I wasn’t going to be the asshole who brings a smoked trout salad to work on a day when I knew I’d need to eat at my desk, in smelling distance of my poor seat mates. ↩
Recipe: 1/4 cup steel-cut oats cooked in 1 cup of water. Stir in a serving of the aforementioned dried tart Montmorency cherries, a tablespoon of almond butter, and a tablespoon of maple syrup. ↩
The importance of low-carb breakfasts to my blood glucose control was one of the key takeaways from my low-carb diet experimentation. ↩
It doesn’t even seem to me that “lot number” is the right differentiator here. I thought the P632/P718 number was a lot number, but it seems to be more of a model number. ↩